- Email: [email protected]
Antibody Response to Burkholderia cepacia in Patients with Cystic Fibrosis Colonized with Burkholderia cepacia and Pseudomonas aeruginosa
Journal of Infection (2000) 40, 164–170 doi:10.1053/jinf.1999.0626, available online at http://www.idealibrary.com on
Antibody Response to Burkholderia cepacia in Patients with Cystic Fibrosis Colonized with Burkholderia cepacia and Pseudomonas aeruginosa J. Hendry*,1, S. Butler3, J. S. Elborn2, J. R. W. Govan3, J. Nelson3, D. J. Shale4 and A. K. Webb1 1 Bradbury Cystic Fibrosis Unit, Wythenshawe Hospital, Manchester, 2Belfast City Hospital, Lisburn Road, Belfast BT9 7AB, U.K., 3Department of Medical Microbiology, Medical School, University of Edinburgh, and 4 University of Wales College of Medicine, Llandough Hospital, Penarth, Cardiff, U.K.
Introduction: This study was designed to determine the relationship between formation of serum antibodies to lipopolysaccharide (LPS) core antigen of Burkholderia cepacia and pulmonary colonization with B. cepacia and Pseudomonas aeruginosa in patients with cystic fibrosis (CF), and to define if an enhanced host humoral immune response to B. cepacia was related to a poor clinical outcome. Methods: Serum IgG to B. cepacia LPS core antigen was measured in adult cystic fibrosis patients colonized with B. cepacia and P. aeruginosa, and serial titres were measured in 13 B. cepacia and 41 P. aeruginosa colonized patients followed prospectively over 18 months. Results: The median B. cepacia antibody titre was significantly greater in the patients colonized with B. cepacia compared to those colonized with P. aeruginosa, a group which grew B. cepacia intermittently from their sputum, and nine healthy controls. The median antibody titre at recruitment into the study was significantly greater in patients who later went into exacerbations compared with those who remained clinically stable, but there was no difference between B. cepacia antibody titres in patients who died and those who survived the study duration. Discussion: The degree of overlap of serum IgG levels to B. cepacia LPS core antigen in cystic fibrosis patients colonized with B. cepacia and P. aeruginosa does not allow this antibody to be used in a clinical context to define infection status. The magnitude of the humoral response to B. cepacia may influence occurrence of pulmonary exacerbations, but a more exuberant humoral immune response to B. cepacia core LPS is not the mechanism by which pulmonary deterioration occurs. © 2000 The British Infection Society
Introduction Pulmonary colonization with Burkholderia cepacia in cystic fibrosis (CF) is associated with a range of clinical outcomes, varying from chronic asymptomatic carriage to acceleration of a gradual decline in respiratory function, and in some cases a fulminant, often fatal pneumonic illness with a high temperature and leukocytosis.1–3 The mechanism by which B. cepacia inflicts pulmonary damage is unclear, particularly as studies to date have failed to identify any convincing virulence factors to account for its pathogenicity in the cystic fibrosis lung.4–6 Furthermore, in animal models B. cepacia appears to be less virulent and to have less capacity for tissue invasion than Pseudomonas aeruginosa.7 In the CF lung colonized
*Address correspondence to: J. Hendry, 27A Moseley Road, Cheadle Humle, Cheshire SK8 5HJ, U.K. Accepted for publication 22 December 1999. 0163-4453/00/020164]07 $35.00/0
with P. aeruginosa, the presence of a chronic bacterial load causes ongoing immune and inflammatory activity within the airways, culminating in the release of neutrophil degranulation products which may cause damage to pulmonary tissues as ‘innocent bystanders’.8–10 Possible mechanisms to explain the pathogenicity of B. cepacia include provocation of a more intense immune and inflammatory response within the airways than P. aeruginosa, mediated through promoting increased neutrophil recruitment and priming of neutrophil burst responses.11–14 Little is known about the humoral immune response to B. cepacia infection in CF. Two studies15,16 detected IgG antibodies to outer membrane proteins of B. cepacia in the sera of patients with CF colonized both with B. cepacia and P. aeruginosa, causing speculation of antigenic crossreactivity between the outer membrane proteins of these organisms. A more recent study17 produced similar results, and confirmed that the antibody response was specific to B. cepacia antigens. An enzyme linked immunosorbant assay © 2000 The British Infection Society
Antibody Response to B. cepacia in Cystic Fibrosis Patients (ELISA) measuring titres of serum IgG and sputum IgA antibodies to B. cepacia lipopolysaccharide (LPS) core antigen has also been developed;18 lack of cross-reactivity between B. cepacia and P. aeruginosa LPS was demonstrated by Western blotting and absorption studies. Using this assay, titres of serum IgG and sputum IgA against B. cepacia LPS core antigen were significantly greater in CF patients colonized with B. cepacia than age- and sexmatched CF patients colonized with P. aeruginosa and healthy individuals without CF and harbouring neither organism. In the CF lung colonized with P. aeruginosa, the presence of antibodies to numerous pseudomonal antigens does not protect the host. For example, despite the presence of serum IgG and sputum IgA antibodies against pseudomonal flagella antigens,19 there is no protective effect against colonization. In chronic P. aeruginosa colonization, serum IgG and sputum IgA antibodies to a variety of P. aeruginosa antigens have been detected, including pooled P. aeruginosa antigens,20–22 P. aeruginosa elastase, alkaline protease23 and LPS,24 and a lag has been demonstrated between the onset of chronic colonization and detection of the antibody response. In one study,25 levels of anti-P. aeruginosa exotoxin A and LPS antibodies were greater in those who died within the study period than those who survived. Conversely, it is recognized that persistent hypogammaglobulinaemia in CF defines a subgroup of patients with better nutritional status, fewer hospital admissions and better preservation of spirometry than patients with normal or hypergammaglobulinaemia.26 High titres of anti-P. aeruginosa antibodies in chronically infected patients tend to have low opsonic activity, and are ineffective at achieving bacterial clearance.28 Additionally, elevated P. aeruginosa antibody titres may promote immune complex formation and enhance the inflammatory response and injury to the lungs. This study utilizes the LPS core antibody assay developed for B. cepacia by Nelson et al.,18 which is able to differentiate between B. cepacia and P. aeruginosa LPS. One aim was to determine the relationship between antibody formation and the presence of chronic pulmonary infection with B. cepacia, P. aeruginosa and intermittent B. cepacia infection. A further objective was to determine the relationship of longterm serum IgG response to B. cepacia to clinical outcome in chronically infected subjects; with the aim of defining if an enhanced host humoral immune response to B. cepacia was related to a poor clinical outcome.
Patients and Methods At the start of the study, 13 patients with CF were colonized with B. cepacia, defined as three successive sputum
165
cultures within a 6-month period. All 13 patients agreed to participate in the study. A further six patients not meeting the criteria for colonization grew B. cepacia in their sputum intermittently, and also agreed to participate in the study. Patients colonized with P. aeruginosa were invited to participate in the study if their clinical severity based upon FEV1% predicted, Shwachman score, and body mass index fell within the range in the B. cepacia colonized group. Forty-one such individuals participated in the study. Approximately 150 patients regularly attended the Manchester Adult Cystic Fibrosis clinic at commencement of the study. In order to determine the pattern of serum IgG antibodies to B. cepacia LPS in the clinic population, those patients not in the prospective study were asked to donate an aliquot of venous blood during the 18-month duration of the study, and 57 samples were obtained. Serum samples were also obtained from nine healthy healthcare professionals working on the CF unit, to provide a control population. Study design For patients in the prospective study, the frequency of assessment was determined by their clinical course. Patients with a pulmonary exacerbation, defined as an increase in pulmonary symptoms and decrease in objective measurements of pulmonary function of sufficient magnitude to require intravenous antimicrobial therapy, were reviewed before commencement of antibiotics and within 48 h of completion of treatment. Intravenous antibiotics were typically administered for 14 days, however, if the patient had continuing respiratory symptoms (cough, breathlessness, increased daily sputum volume) and or suboptimal spirometry compared to their normal baseline values; antimicrobial therapy was extended until resolution had occurred. Clinically stable patients with no deterioration in spirometry, Shwachman score or body mass over a period of 12 months preceding the study were reviewed at 3-monthly intervals. At each attendance, 10 ml of venous blood were drawn and were allowed to clot at room temperature before separation by centrifugation at 2000 rpm for 10 min. Serum was stored in polypropylene tubes at [70 ºC until analysis. Serum IgG to B. cepacia LPS core antigen was measured by ELISA.18 Extracted LPS from a rough B. cepacia isolate (J1359) was used to coat microtitre plates. Extracted LPS was complexed with polymyxin and used at a final concentration of 10 ng/ml. LPS-polymyxin complexes were diluted in coating buffer and added to microtitration plates at 100 mg/well. All plates were coated overnight at room temperature and washed four times
J. Hendry et al.
166
with wash buffer (PBS, pH 7.2, containing Tween 20 0.05%w/v and sodium azide 0.02%w/v). All plates were then post-coated with post-coat buffer consisting of PBS containing bovine serum albumin (BSA) 5%w/v at 100 ml per well. After being washed four times with wash buffer, plates were stored at [20 ºC until used. Serum specimens were serially diluted in buffer and added to coated microtitre plates in triplicate in 100 ml aliquots. The plates were incubated at 37 ºC for 90 min, then washed four times with wash buffer. Conjugates, including alkaline phosphatase conjugated to antihuman IgG (Sigma), were added to each well in 100 ml aliquots. Plates were incubated for a further 90 min at 37 ºC, washed and rinsed before addition of 100 ml of alkaline phosphatase substrate (Sigma) to each well. The plates were incubated at room temperature for 30 min before the optical density was read at 405 nm in a Titertek Multiscan plate reader (Flow Laboratories Ltd, Irvine, U.K). Final results were expressed as the titre giving OD[0.1 after subtraction of the OD of negative controls for each sample. The intra and interplate variations of the ELISA were 5.5% and 13.9%, respectively.
for culture of Staphylococcus aureus. Routine antibiotic sensitivity testing was carried out using the modified Stokes technique. Isolates of B. cepacia were confirmed using the API 20NE multitest system and ‘fingerprinted’ by pulsed field gel electrophoresis. Statistical methods Antibody titres were not normally distributed, therefore results are presented as the median (range). Comparisons between groups were made using non-parametric statistical tests. Statistical significance was accepted when P\0.05.
Results Prospective study Ages and indicators of clinical severity in the three groups at the time of recruitment into the prospective study were well matched (Table I). Sputum microbiology. Results are summarized in Table II.
Microbiological methods At each visit, sputum was obtained for culture and antibiotic sensitivity. Sputum was plated onto CLED medium, blood agar and B. cepacia medium (Mast cepacia agar; Mast Diagnostics Ltd, Bootle, U.K.). Pseudomonas aeruginosa was identified by colonial morphology and production of pyocyanin on Pseudomonas isolation agar (Difco Labs. Detroit, Michigan, U.S.A.); non-pigmented or atypical isolates were confirmed by a positive oxidase test and a characteristic biochemical profile using API 20NE multitest system (API-bioMerieux, MarcylEtoile, France). Sputum was also plated onto Bacitracin chocolate agar for culture of Haemophilus influenzae, and Mannitol salt agar
Changes in B. cepacia antibody levels over time in the three groups. For each patient prospective study, the B. cepacia antibody titre was compared between the first and final blood samples. The median increase in B. cepacia antibody titres was significantly greater in patients colonized with B. cepacia (39806 (range [28951 to 119236)) compared to the median increase in those colonized with P. aeruginosa (1282 (range [1440 to 7629)), P\0.001, but not compared with patients intermittently colonized with B. cepacia (8040 ([6574 to 29163)), P[0.05. There was no difference between the increase in B. cepacia antibody titres in the group colonized with P. aeruginosa and the intermittent growers of B. cepacia, P[0.05.
Table I. Comparison of clinical parameters within the three study groups (B. cepacia, P. aeruginosa, B. cepacia intermittent) at the time of recruitment to the study.
Age median (range) FEV1% pre mean (SD) Body mass index mean (SD) Shwachman score median (range)
B. cepacia
P. aeruginosa
B. cepacia intermittent
Test statistic
20.1 (17.1, 30.2)
20.6 (16.9, 46.1)
26.4 (18.3, 36.4)
P[0.05
54.4 (22.0)
59.1 (26.8)
54.3 (28.6)
P[0.05
19.6 (2.7)
19.6 (1.8)
20.2 (2.7)
P[0.05
66 (38, 74)
66 (50, 88)
67 (59, 77)
P[0.05
Antibody Response to B. cepacia in Cystic Fibrosis Patients
167
Table II. Sputum microbiology in the groups colonized with P. aeruginosa and B. cepacia.
Patients colonized with B. cepacia (n\13)
Patients colonized with P. aeruginosa (n\41)*
B. cepacia alone
B. cepacia and S. aureus
B. cepacia and P. aeruginosa
B. cepacia and P. aeruginosa and S. aureus
3
1
6
3
P. aeruginosa alone
P. aeruginosa and S. aureus
P. aeruginosa and S. aureus and S. maltophilia
P. aeruginosa and S. maltophilia
20
17
2
1
* In a single patient no bacterial pathogens were isolated during the study period, although this patient had previously been colonized with P. aeruginosa.
A comparison of B. cepacia antibody levels in patients in continuous clinical stability and those with pulmonary exacerbations. In the 13 patients colonized with B. cepacia, 11 had a pulmonary exacerbation during the study, and two remained in continuous clinical stability. The median B. cepacia antibody titre at recruitment into the study was significantly greater in patients who later went into exacerbations (5023 (range 2900–64500)) compared with those in continuous clinical stability (1251 (range 238–2264)), P\0.05. Forty-one patients colonized with P. aeruginosa but culture-negative for B. cepacia experienced an exacerbation during the study, and six remained clinically stable. The median B. cepacia antibody level was significantly greater in the group who later had exacerbations (579 (range 0–4917)) compared with continuously stable patients (200 (range 0–314)), P\0.05. Of the group with intermittent isolation of B. cepacia, only one patient remained in continuous clinical stability, so no comparison within that group was possible. A comparison of B. cepacia antibodies in survivors and nonsurvivors. Of the 13 patients colonized with B. cepacia, four died during the study. There was no difference between B. cepacia antibody titres between patients who died and those who survived, using values either at the closest date to recruitment, or the closest date to exit from the study; P[0.05. Similarly, among patients colonized with P. aeruginosa (repeatedly culture-negative; for B. cepacia), there was no difference between B. cepacia antibody titres in patients who died or those who survived using the same assessment points (P[0.05). There were no deaths
among patients in the group who grew B. cepacia from their sputum intermittently. The relationship between antibodies to B. cepacia core LPS and the strain of B. cepacia. Pulsed-field gel electrophoresis patterns were obtained for 12 of the 13 patients colonized with B. cepacia. One patient died before sputum was obtained for this purpose. At first measurement within the study, there was no difference between B. cepacia core LPS antibody levels in the five groups representing different strains of B. cepacia, P[0.05.
Cross-sectional study of serum IgG to B. cepacia core LPS For the 57 patients who were included in the crosssectional review of the clinic population (but not the prospective study), sputum isolates were as follows: 18 P. aeruginosa alone, 16 P. aeruginosa and S. aureus, six P. aeruginosa and B. cepacia, three B. cepacia alone, one P. aeruginosa and B. cepacia and S. aureus, six S. aureus alone, five normal flora, and two patients were unable to expectorate sputum. All 10 patients who were culturepositive for B. cepacia were colonized according to our criteria. These patients were amalgamated with those studied prospectively to form three groups: Group 1, those patients colonized by P. aeruginosa, S. aureus and culturenegative for B. cepacia, (n\88); Group 2, those patients colonized with B. cepacia, with or without other organisms (n\23); Group 3, those patients isolating B. cepacia from sputum intermittently (n\6).
168
Figure 1.
J. Hendry et al.
Serum IgG to B. cepacia core LPS in patients with CF colonized with P. aeruginosa, B. cepacia and intermittent growers of B. cepacia.
The median B. cepacia antibody titre was significantly greater in the patients colonized with B. cepacia (5023 (range 238–72663) EU) compared to those colonized with P. aeruginosa (289 (range 0–4917) EU), the group which grew B. cepacia intermittently from their sputum (1915 (range 200–5973) EU), and nine healthy controls (100 (range (100–2400) EU), P\0.001 (Fig. 1). The median B. cepacia antibody titres were greater in intermittent growers of B. cepacia compared with healthy controls (P\0.05), but not compared with those colonized with P. aeruginosa, (P[0.05). Changes in B. cepacia antibody titres in patients colonized with P. aeruginosa who subsequently became colonized with B. cepacia. Six of the 47 patients colonized with P. aeruginosa in the cross-sectional study became colonized with B. cepacia between the end of the study up to 1997. During the prospective study, eight of the 41 patients colonized with P. aeruginosa became colonized with B. cepacia and five after completion of the study. The final B. cepacia antibody titre for each patient in the prospective study was analysed. For the patients who became colonized within the study, the last B. cepacia antibody titre before the first isolation of B. cepacia has been used in the analysis. There was no difference between the median B. cepacia antibody titre in the group who became colonized with B. cepacia during the study, pooled with the group who became colonized between the end of the study and the present day
(1654 (range 100–8028)) compared to the group who remained free from B. cepacia colonization (1051 (range 0–8000)), P[0.05.
Discussion The cross-sectional analysis of B. cepacia antibody titres confirmed that titres of serum antibodies to the core LPS of B. cepacia were greater in CF patients colonized by B. cepacia than those colonized by P. aeruginosa, or those who grow B. cepacia intermittently from their sputum. However, titres overlapped between the groups, making it impossible to classify patients as B. cepacia colonized without knowledge of their sputum microbiology. Within the B. cepacia colonized patients, there were no differences in antibody titres when the group was subdivided by strain, implying that the antibody has a broad range of specificity for an antigen conserved between different strains of B. cepacia. The rate of rise of B. cepacia antibody titres with time varied among individual patients colonized with B. cepacia, but the overall trend was for an increase with time. Antibody titres in those who became colonized with B. cepacia during the study were not different to those colonized after the end of the study, P[0.05. This suggests that although B. cepacia antibody levels may rise before the first isolation of B. cepacia from sputum, this is not invariably the case. Some CF patients colonized with
Antibody Response to B. cepacia in Cystic Fibrosis Patients P. aeruginosa may have B. cepacia antibody titres in their serum which are of similar magnitude to those measured in the serum of patients who subsequently become B. cepacia colonized over the course of weeks or months, and, furthermore, those titres may overlap values observed in B. cepacia colonized patients and intermittent growers. In patients colonized with B. cepacia, those who experienced exacerbations had greater titres compared with those who remained in clinically stable. This suggests that the magnitude of the humoral response to B. cepacia may influence occurrence of exacerbations. However, no difference was demonstrated between B. cepacia antibody titres in colonized patients who died during the study and those who survived. Hence, serum antibody responses to B. cepacia core LPS as determined here do not support the speculation that a more exuberant humoral immune response to B. cepacia may be the mechanism by which pulmonary deterioration occurs. However, this interpretation must be tempered by the limitations of current knowledge about the host immune response to B. cepacia. Studies in patients with CF who are colonized with P. aeruginosa suggest that IgG antibodies afford no protection against infection with P. aeruginosa, fail to clear the organism, and, paradoxically, may participate in and perpetuate an immune and inflammatory response leading to pulmonary injury.25–27 Hence, a high antibody titre and formation of immune complexes are poor prognostic features.28 Our studies provide no evidence that serum IgG B. cepacia antibodies have the same significance for lung injury in those colonized by B. cepacia. These antibodies do not appear to achieve clearance of B. cepacia from the lung, as evidenced by continuing colonization in the face of increasing antibody titres. However, our study does not support the hypothesis that high antibody titres are associated with poor prognosis. This conclusion relates only to the particular antibody utilized in the study. It is possible that antibodies to different antigenic components of B. cepacia, including flagellar and Oside chain antigen, may have different pathophysiological significance. While it is clear that production of IgG antibodies against B. cepacia LPS does not clear the organism from the lungs, it is not clear whether the presence of this antibody in the serum of patients colonized with P. aeruginosa may afford future protection against pulmonary colonization with B. cepacia. This study did not indicate any difference in B. cepacia LPS antibody titres between P. aeruginosa colonized patients who subsequently became colonized with B. cepacia, and therefore suggests no protective interaction. The division of B. cepacia patients into those chronically colonized and those who grew the organism intermittently
169
was based upon the definition of colonization without taking into account differences in clinical behaviour. Greater B. cepacia antibody titres in the colonized group indicates a heightened humoral immune response to the organism compared to the intermittent growers, though antibody levels in the intermittent growers of B. cepacia were greater than in those colonized with P. aeruginosa, P\0.05. An alternative explanation for our findings is that the antibody response to B. cepacia core antigen is related to the ‘burden’ or duration of infection, and may be only partially related to the pathophysiology of lung injury. In this respect, these findings are similar to those relating to antibody patterns in chronic pulmonary infection with P. aeruginosa in CF. The degree of overlap of antibody titres demonstrated here does not allow antibody levels to be used in a clinical context to define infection status in patients with CF.
References 01. Isles A, Macluskey, I, Corey M. Pseudomonas cepacia infection in cystic fibrosis: an emerging problem. J Paediatr 1984; 104: 206–210. 02. Thomassen MJ, Demko CA, Klinger JD, Stern RC. Pseudomonas cepacia: colonisation among patients with cystic fibrosis. Am Rev Resp Dis 1985; 131: 791–796. 03. Simmonds EJ, Conway SP, Ghonheim ATM, Littlewood JM. Pseudomonas cepacia: a new pathogen in patients with cystic fibrosis referred to a large centre in the United Kingdom. Arch Dis Child 1990; 65: 874–877. 04. Gessner AR, Mortensen JE. Pathogenic factors of Pseudomonas cepacia isolates from patients with Cystic Fibrosis. J Med Microbiol 1990; 33: 115–120. 05. Govan JRW, Hughes JE, Vandamme P. Burkholderia cepacia: medical, taxonomic and ecological issues. J Med Microbiol 1996; 45: 385–407. 06. Nelson JW, Butler SL, Kreig D, Govan JRW. Virulence factors of Burkholderia cepacia. FEMS Immunol Med Microbiol 1994; 8: 89–97. 07. Stover G, Drake DR, Montie TC. Virulence of different Pseudomonas species in a burned mouse model: tissue colonisation by Pseudomonas cepacia. Infect Immun 1983; 41: 1099–1144. 08. Suter S, Schaad UB, Roux-Lombard P, Girardin E, Grau G, Dayer JM. Relation between TNF-a and granulocyte elastase-a1-proteinase inhibitor complexes in plasma of patients with cystic fibrosis. Am Rev Resp Dis 1989; 140: 1640–1644. 09. Suter S, Schaad Uh, Roux L, Nydegger U, Waldvogel F. Granulocyte neutral proteases and pseudomonal elastase as possible causes of airway damage in patients with cystic fibrosis. J Infect Dis 1984; 149: 523–531. 10. Meyer K, Lewandowski J, Zimmerman J, Nunley D, Calhoun W, Dopico G. Human neutrophil elastase a-1-antiprotease complex in cystic fibrosis. Am Rev Resp Dis 1991; 144: 580–585. 11. Hughes JE, Stewart J, Barclay GR, Govan JRW. Priming of neutrophil burst activity by lipopolysaccharide of Burkholderia cepacia. Infect Immun 1997; 65: 4281–4287. 12. Hutchinson ML, Poxton IR, Govan JRW. Burkholderia cepacia produces a haemolysin that is capable of inducing apoptosis and degranulation of mammalian phagocytes. Infect Immun 1998; 66: 2033–2039. 13. Zughaier SM, Ryley HC, Jackson SK. A melanin pigment purified from an epidemic strain of Burkholderia cepacia attenuates monocyte respiratory burst activity by scavenging superoxide anion. Infect Immun 1999; 67: 908–913.
170
J. Hendry et al.
14. Zughaier SM, Ryley HC, Jackson SK. Lipopolysaccharide (LPS) from Burkholderia cepacia is more active than LPS from Pseudomonas aeruginosa and Stenotrophomonas maltophilia in stimulating tumour necrosis factor alpha from human monocytes. Infect Immun 1999; 67: 1505–1507. 15. Aronoff SC, Quiner FJ, Stern RC. Longitudinal serum IgG response to Pseudomonas cepacia surface antigens in cystic fibrosis. Paediatr Pulmonol 1991; 1: 289–293. 16. Aronoff SC, Stern RC. Serum IgG antibody to the outer membrane antigens of Pseudomonas cepacia and Pseudomonas aeruginosa in cystic fibrosis. J Infect Dis 1988; 157: 934–940. 17. Lacy D, Smith A, Stableforth D, Smith G, Weller P, Brown M. Serum IgG response to B. cepacia outer membrane antigens in cystic fibrosis: assessment of cross-reactivity with P. aeruginosa. FEMS Immunology and Medical Microbiology 1995; 10: 253–262. 18. Nelson JW, Butler SL, Brown PH, Greening AP, Govan JRW. Serum IgG and sputum IgA antibody to core lipopolysaccharide antigen from Pseudomonas cepacia in patients with cystic fibrosis. J Med Microbiol 1993; 39: 39–47. 19. Nelson JW, Govan JRW, Barclay GR. Pseudomonas aeruginosa flagellar antibodies in serum, saliva and sputum from patients with cystic fibrosis. Serodiag Immunother Inf Dis 1990; 4: 351–361. 20. Brett MM, Ghonheim ATM, Littlewood JM. Serum IgG antibodies in patients with cystic fibrosis with early P. aeruginosa infection. Arch Dis Child 1987; 62: 357–361. 21. Doring G, Hoiby N. Longitudinal study of immune response to Pseudomonas aeruginosa antigens in cystic fibrosis. Infect Immun 1983; 42: 197–201.
22. Nelson JW, Butler SL, Brown PH, Greening AP, Govan JRW. Serum IgG and sputum IgA antibody to core lipopolysaccharide antigen from Pseudomonas cepacia in patients with cystic fibrosis. J Med Microbiol 1993; 39: 39–47. 23. Hollsing AE, Granstrom M, Vasil ML et al. Prospective study of serum antibodies to Pseudomonas aeruginosa exoproteins in cystic fibrosis. J Clin Microbiol 1987; 25: 1830–1836. 24. Nelson JW, Barclay R, Govan JRW. Diagnosis of chronic Pseudomonas aeruginosa infection in cystic fibrosis by enzyme linked immunosorbent assay for anti pseudomonas LPS antibodies. Serodiag Immunother Infect Dis 1990; 4: 9–16. 25. Moss RB. Hypergammaglobulinaemia in cystic fibrosis. Chest 1987; 91: 522–526. 26. Wheeler WB, Williams M, Matthews WJ, Colten HR. Progression of cystic fibrosis lung disease as a function of serum immunoglobulin levels: a five year longitudinal study. J Paediatr 1984; 104: 695– 699. 27. Tosi M, Zakem-Cloud H, Demko C et al. Cross-sectional and longitudinal studies of naturally occuring antibodies to Pseudomonas aeruginosa in cystic fibrosis. J Inf Dis 1995; 172: 453–461. 28. Manthei U, Taussig LM, Beckerman RC, Strunk RC. Circulating immune complexes in cystic fibrosis. Am Rev Resp Dis 1982; 126: 253–257.
Antibody Response to Burkholderia cepacia in Patients with Cystic Fibrosis Colonized with Burkholderia cepacia and Pseudomonas aeruginosa J. Hendry*,1, S. Butler3, J. S. Elborn2, J. R. W. Govan3, J. Nelson3, D. J. Shale4 and A. K. Webb1 1 Bradbury Cystic Fibrosis Unit, Wythenshawe Hospital, Manchester, 2Belfast City Hospital, Lisburn Road, Belfast BT9 7AB, U.K., 3Department of Medical Microbiology, Medical School, University of Edinburgh, and 4 University of Wales College of Medicine, Llandough Hospital, Penarth, Cardiff, U.K.
Introduction: This study was designed to determine the relationship between formation of serum antibodies to lipopolysaccharide (LPS) core antigen of Burkholderia cepacia and pulmonary colonization with B. cepacia and Pseudomonas aeruginosa in patients with cystic fibrosis (CF), and to define if an enhanced host humoral immune response to B. cepacia was related to a poor clinical outcome. Methods: Serum IgG to B. cepacia LPS core antigen was measured in adult cystic fibrosis patients colonized with B. cepacia and P. aeruginosa, and serial titres were measured in 13 B. cepacia and 41 P. aeruginosa colonized patients followed prospectively over 18 months. Results: The median B. cepacia antibody titre was significantly greater in the patients colonized with B. cepacia compared to those colonized with P. aeruginosa, a group which grew B. cepacia intermittently from their sputum, and nine healthy controls. The median antibody titre at recruitment into the study was significantly greater in patients who later went into exacerbations compared with those who remained clinically stable, but there was no difference between B. cepacia antibody titres in patients who died and those who survived the study duration. Discussion: The degree of overlap of serum IgG levels to B. cepacia LPS core antigen in cystic fibrosis patients colonized with B. cepacia and P. aeruginosa does not allow this antibody to be used in a clinical context to define infection status. The magnitude of the humoral response to B. cepacia may influence occurrence of pulmonary exacerbations, but a more exuberant humoral immune response to B. cepacia core LPS is not the mechanism by which pulmonary deterioration occurs. © 2000 The British Infection Society
Introduction Pulmonary colonization with Burkholderia cepacia in cystic fibrosis (CF) is associated with a range of clinical outcomes, varying from chronic asymptomatic carriage to acceleration of a gradual decline in respiratory function, and in some cases a fulminant, often fatal pneumonic illness with a high temperature and leukocytosis.1–3 The mechanism by which B. cepacia inflicts pulmonary damage is unclear, particularly as studies to date have failed to identify any convincing virulence factors to account for its pathogenicity in the cystic fibrosis lung.4–6 Furthermore, in animal models B. cepacia appears to be less virulent and to have less capacity for tissue invasion than Pseudomonas aeruginosa.7 In the CF lung colonized
*Address correspondence to: J. Hendry, 27A Moseley Road, Cheadle Humle, Cheshire SK8 5HJ, U.K. Accepted for publication 22 December 1999. 0163-4453/00/020164]07 $35.00/0
with P. aeruginosa, the presence of a chronic bacterial load causes ongoing immune and inflammatory activity within the airways, culminating in the release of neutrophil degranulation products which may cause damage to pulmonary tissues as ‘innocent bystanders’.8–10 Possible mechanisms to explain the pathogenicity of B. cepacia include provocation of a more intense immune and inflammatory response within the airways than P. aeruginosa, mediated through promoting increased neutrophil recruitment and priming of neutrophil burst responses.11–14 Little is known about the humoral immune response to B. cepacia infection in CF. Two studies15,16 detected IgG antibodies to outer membrane proteins of B. cepacia in the sera of patients with CF colonized both with B. cepacia and P. aeruginosa, causing speculation of antigenic crossreactivity between the outer membrane proteins of these organisms. A more recent study17 produced similar results, and confirmed that the antibody response was specific to B. cepacia antigens. An enzyme linked immunosorbant assay © 2000 The British Infection Society
Antibody Response to B. cepacia in Cystic Fibrosis Patients (ELISA) measuring titres of serum IgG and sputum IgA antibodies to B. cepacia lipopolysaccharide (LPS) core antigen has also been developed;18 lack of cross-reactivity between B. cepacia and P. aeruginosa LPS was demonstrated by Western blotting and absorption studies. Using this assay, titres of serum IgG and sputum IgA against B. cepacia LPS core antigen were significantly greater in CF patients colonized with B. cepacia than age- and sexmatched CF patients colonized with P. aeruginosa and healthy individuals without CF and harbouring neither organism. In the CF lung colonized with P. aeruginosa, the presence of antibodies to numerous pseudomonal antigens does not protect the host. For example, despite the presence of serum IgG and sputum IgA antibodies against pseudomonal flagella antigens,19 there is no protective effect against colonization. In chronic P. aeruginosa colonization, serum IgG and sputum IgA antibodies to a variety of P. aeruginosa antigens have been detected, including pooled P. aeruginosa antigens,20–22 P. aeruginosa elastase, alkaline protease23 and LPS,24 and a lag has been demonstrated between the onset of chronic colonization and detection of the antibody response. In one study,25 levels of anti-P. aeruginosa exotoxin A and LPS antibodies were greater in those who died within the study period than those who survived. Conversely, it is recognized that persistent hypogammaglobulinaemia in CF defines a subgroup of patients with better nutritional status, fewer hospital admissions and better preservation of spirometry than patients with normal or hypergammaglobulinaemia.26 High titres of anti-P. aeruginosa antibodies in chronically infected patients tend to have low opsonic activity, and are ineffective at achieving bacterial clearance.28 Additionally, elevated P. aeruginosa antibody titres may promote immune complex formation and enhance the inflammatory response and injury to the lungs. This study utilizes the LPS core antibody assay developed for B. cepacia by Nelson et al.,18 which is able to differentiate between B. cepacia and P. aeruginosa LPS. One aim was to determine the relationship between antibody formation and the presence of chronic pulmonary infection with B. cepacia, P. aeruginosa and intermittent B. cepacia infection. A further objective was to determine the relationship of longterm serum IgG response to B. cepacia to clinical outcome in chronically infected subjects; with the aim of defining if an enhanced host humoral immune response to B. cepacia was related to a poor clinical outcome.
Patients and Methods At the start of the study, 13 patients with CF were colonized with B. cepacia, defined as three successive sputum
165
cultures within a 6-month period. All 13 patients agreed to participate in the study. A further six patients not meeting the criteria for colonization grew B. cepacia in their sputum intermittently, and also agreed to participate in the study. Patients colonized with P. aeruginosa were invited to participate in the study if their clinical severity based upon FEV1% predicted, Shwachman score, and body mass index fell within the range in the B. cepacia colonized group. Forty-one such individuals participated in the study. Approximately 150 patients regularly attended the Manchester Adult Cystic Fibrosis clinic at commencement of the study. In order to determine the pattern of serum IgG antibodies to B. cepacia LPS in the clinic population, those patients not in the prospective study were asked to donate an aliquot of venous blood during the 18-month duration of the study, and 57 samples were obtained. Serum samples were also obtained from nine healthy healthcare professionals working on the CF unit, to provide a control population. Study design For patients in the prospective study, the frequency of assessment was determined by their clinical course. Patients with a pulmonary exacerbation, defined as an increase in pulmonary symptoms and decrease in objective measurements of pulmonary function of sufficient magnitude to require intravenous antimicrobial therapy, were reviewed before commencement of antibiotics and within 48 h of completion of treatment. Intravenous antibiotics were typically administered for 14 days, however, if the patient had continuing respiratory symptoms (cough, breathlessness, increased daily sputum volume) and or suboptimal spirometry compared to their normal baseline values; antimicrobial therapy was extended until resolution had occurred. Clinically stable patients with no deterioration in spirometry, Shwachman score or body mass over a period of 12 months preceding the study were reviewed at 3-monthly intervals. At each attendance, 10 ml of venous blood were drawn and were allowed to clot at room temperature before separation by centrifugation at 2000 rpm for 10 min. Serum was stored in polypropylene tubes at [70 ºC until analysis. Serum IgG to B. cepacia LPS core antigen was measured by ELISA.18 Extracted LPS from a rough B. cepacia isolate (J1359) was used to coat microtitre plates. Extracted LPS was complexed with polymyxin and used at a final concentration of 10 ng/ml. LPS-polymyxin complexes were diluted in coating buffer and added to microtitration plates at 100 mg/well. All plates were coated overnight at room temperature and washed four times
J. Hendry et al.
166
with wash buffer (PBS, pH 7.2, containing Tween 20 0.05%w/v and sodium azide 0.02%w/v). All plates were then post-coated with post-coat buffer consisting of PBS containing bovine serum albumin (BSA) 5%w/v at 100 ml per well. After being washed four times with wash buffer, plates were stored at [20 ºC until used. Serum specimens were serially diluted in buffer and added to coated microtitre plates in triplicate in 100 ml aliquots. The plates were incubated at 37 ºC for 90 min, then washed four times with wash buffer. Conjugates, including alkaline phosphatase conjugated to antihuman IgG (Sigma), were added to each well in 100 ml aliquots. Plates were incubated for a further 90 min at 37 ºC, washed and rinsed before addition of 100 ml of alkaline phosphatase substrate (Sigma) to each well. The plates were incubated at room temperature for 30 min before the optical density was read at 405 nm in a Titertek Multiscan plate reader (Flow Laboratories Ltd, Irvine, U.K). Final results were expressed as the titre giving OD[0.1 after subtraction of the OD of negative controls for each sample. The intra and interplate variations of the ELISA were 5.5% and 13.9%, respectively.
for culture of Staphylococcus aureus. Routine antibiotic sensitivity testing was carried out using the modified Stokes technique. Isolates of B. cepacia were confirmed using the API 20NE multitest system and ‘fingerprinted’ by pulsed field gel electrophoresis. Statistical methods Antibody titres were not normally distributed, therefore results are presented as the median (range). Comparisons between groups were made using non-parametric statistical tests. Statistical significance was accepted when P\0.05.
Results Prospective study Ages and indicators of clinical severity in the three groups at the time of recruitment into the prospective study were well matched (Table I). Sputum microbiology. Results are summarized in Table II.
Microbiological methods At each visit, sputum was obtained for culture and antibiotic sensitivity. Sputum was plated onto CLED medium, blood agar and B. cepacia medium (Mast cepacia agar; Mast Diagnostics Ltd, Bootle, U.K.). Pseudomonas aeruginosa was identified by colonial morphology and production of pyocyanin on Pseudomonas isolation agar (Difco Labs. Detroit, Michigan, U.S.A.); non-pigmented or atypical isolates were confirmed by a positive oxidase test and a characteristic biochemical profile using API 20NE multitest system (API-bioMerieux, MarcylEtoile, France). Sputum was also plated onto Bacitracin chocolate agar for culture of Haemophilus influenzae, and Mannitol salt agar
Changes in B. cepacia antibody levels over time in the three groups. For each patient prospective study, the B. cepacia antibody titre was compared between the first and final blood samples. The median increase in B. cepacia antibody titres was significantly greater in patients colonized with B. cepacia (39806 (range [28951 to 119236)) compared to the median increase in those colonized with P. aeruginosa (1282 (range [1440 to 7629)), P\0.001, but not compared with patients intermittently colonized with B. cepacia (8040 ([6574 to 29163)), P[0.05. There was no difference between the increase in B. cepacia antibody titres in the group colonized with P. aeruginosa and the intermittent growers of B. cepacia, P[0.05.
Table I. Comparison of clinical parameters within the three study groups (B. cepacia, P. aeruginosa, B. cepacia intermittent) at the time of recruitment to the study.
Age median (range) FEV1% pre mean (SD) Body mass index mean (SD) Shwachman score median (range)
B. cepacia
P. aeruginosa
B. cepacia intermittent
Test statistic
20.1 (17.1, 30.2)
20.6 (16.9, 46.1)
26.4 (18.3, 36.4)
P[0.05
54.4 (22.0)
59.1 (26.8)
54.3 (28.6)
P[0.05
19.6 (2.7)
19.6 (1.8)
20.2 (2.7)
P[0.05
66 (38, 74)
66 (50, 88)
67 (59, 77)
P[0.05
Antibody Response to B. cepacia in Cystic Fibrosis Patients
167
Table II. Sputum microbiology in the groups colonized with P. aeruginosa and B. cepacia.
Patients colonized with B. cepacia (n\13)
Patients colonized with P. aeruginosa (n\41)*
B. cepacia alone
B. cepacia and S. aureus
B. cepacia and P. aeruginosa
B. cepacia and P. aeruginosa and S. aureus
3
1
6
3
P. aeruginosa alone
P. aeruginosa and S. aureus
P. aeruginosa and S. aureus and S. maltophilia
P. aeruginosa and S. maltophilia
20
17
2
1
* In a single patient no bacterial pathogens were isolated during the study period, although this patient had previously been colonized with P. aeruginosa.
A comparison of B. cepacia antibody levels in patients in continuous clinical stability and those with pulmonary exacerbations. In the 13 patients colonized with B. cepacia, 11 had a pulmonary exacerbation during the study, and two remained in continuous clinical stability. The median B. cepacia antibody titre at recruitment into the study was significantly greater in patients who later went into exacerbations (5023 (range 2900–64500)) compared with those in continuous clinical stability (1251 (range 238–2264)), P\0.05. Forty-one patients colonized with P. aeruginosa but culture-negative for B. cepacia experienced an exacerbation during the study, and six remained clinically stable. The median B. cepacia antibody level was significantly greater in the group who later had exacerbations (579 (range 0–4917)) compared with continuously stable patients (200 (range 0–314)), P\0.05. Of the group with intermittent isolation of B. cepacia, only one patient remained in continuous clinical stability, so no comparison within that group was possible. A comparison of B. cepacia antibodies in survivors and nonsurvivors. Of the 13 patients colonized with B. cepacia, four died during the study. There was no difference between B. cepacia antibody titres between patients who died and those who survived, using values either at the closest date to recruitment, or the closest date to exit from the study; P[0.05. Similarly, among patients colonized with P. aeruginosa (repeatedly culture-negative; for B. cepacia), there was no difference between B. cepacia antibody titres in patients who died or those who survived using the same assessment points (P[0.05). There were no deaths
among patients in the group who grew B. cepacia from their sputum intermittently. The relationship between antibodies to B. cepacia core LPS and the strain of B. cepacia. Pulsed-field gel electrophoresis patterns were obtained for 12 of the 13 patients colonized with B. cepacia. One patient died before sputum was obtained for this purpose. At first measurement within the study, there was no difference between B. cepacia core LPS antibody levels in the five groups representing different strains of B. cepacia, P[0.05.
Cross-sectional study of serum IgG to B. cepacia core LPS For the 57 patients who were included in the crosssectional review of the clinic population (but not the prospective study), sputum isolates were as follows: 18 P. aeruginosa alone, 16 P. aeruginosa and S. aureus, six P. aeruginosa and B. cepacia, three B. cepacia alone, one P. aeruginosa and B. cepacia and S. aureus, six S. aureus alone, five normal flora, and two patients were unable to expectorate sputum. All 10 patients who were culturepositive for B. cepacia were colonized according to our criteria. These patients were amalgamated with those studied prospectively to form three groups: Group 1, those patients colonized by P. aeruginosa, S. aureus and culturenegative for B. cepacia, (n\88); Group 2, those patients colonized with B. cepacia, with or without other organisms (n\23); Group 3, those patients isolating B. cepacia from sputum intermittently (n\6).
168
Figure 1.
J. Hendry et al.
Serum IgG to B. cepacia core LPS in patients with CF colonized with P. aeruginosa, B. cepacia and intermittent growers of B. cepacia.
The median B. cepacia antibody titre was significantly greater in the patients colonized with B. cepacia (5023 (range 238–72663) EU) compared to those colonized with P. aeruginosa (289 (range 0–4917) EU), the group which grew B. cepacia intermittently from their sputum (1915 (range 200–5973) EU), and nine healthy controls (100 (range (100–2400) EU), P\0.001 (Fig. 1). The median B. cepacia antibody titres were greater in intermittent growers of B. cepacia compared with healthy controls (P\0.05), but not compared with those colonized with P. aeruginosa, (P[0.05). Changes in B. cepacia antibody titres in patients colonized with P. aeruginosa who subsequently became colonized with B. cepacia. Six of the 47 patients colonized with P. aeruginosa in the cross-sectional study became colonized with B. cepacia between the end of the study up to 1997. During the prospective study, eight of the 41 patients colonized with P. aeruginosa became colonized with B. cepacia and five after completion of the study. The final B. cepacia antibody titre for each patient in the prospective study was analysed. For the patients who became colonized within the study, the last B. cepacia antibody titre before the first isolation of B. cepacia has been used in the analysis. There was no difference between the median B. cepacia antibody titre in the group who became colonized with B. cepacia during the study, pooled with the group who became colonized between the end of the study and the present day
(1654 (range 100–8028)) compared to the group who remained free from B. cepacia colonization (1051 (range 0–8000)), P[0.05.
Discussion The cross-sectional analysis of B. cepacia antibody titres confirmed that titres of serum antibodies to the core LPS of B. cepacia were greater in CF patients colonized by B. cepacia than those colonized by P. aeruginosa, or those who grow B. cepacia intermittently from their sputum. However, titres overlapped between the groups, making it impossible to classify patients as B. cepacia colonized without knowledge of their sputum microbiology. Within the B. cepacia colonized patients, there were no differences in antibody titres when the group was subdivided by strain, implying that the antibody has a broad range of specificity for an antigen conserved between different strains of B. cepacia. The rate of rise of B. cepacia antibody titres with time varied among individual patients colonized with B. cepacia, but the overall trend was for an increase with time. Antibody titres in those who became colonized with B. cepacia during the study were not different to those colonized after the end of the study, P[0.05. This suggests that although B. cepacia antibody levels may rise before the first isolation of B. cepacia from sputum, this is not invariably the case. Some CF patients colonized with
Antibody Response to B. cepacia in Cystic Fibrosis Patients P. aeruginosa may have B. cepacia antibody titres in their serum which are of similar magnitude to those measured in the serum of patients who subsequently become B. cepacia colonized over the course of weeks or months, and, furthermore, those titres may overlap values observed in B. cepacia colonized patients and intermittent growers. In patients colonized with B. cepacia, those who experienced exacerbations had greater titres compared with those who remained in clinically stable. This suggests that the magnitude of the humoral response to B. cepacia may influence occurrence of exacerbations. However, no difference was demonstrated between B. cepacia antibody titres in colonized patients who died during the study and those who survived. Hence, serum antibody responses to B. cepacia core LPS as determined here do not support the speculation that a more exuberant humoral immune response to B. cepacia may be the mechanism by which pulmonary deterioration occurs. However, this interpretation must be tempered by the limitations of current knowledge about the host immune response to B. cepacia. Studies in patients with CF who are colonized with P. aeruginosa suggest that IgG antibodies afford no protection against infection with P. aeruginosa, fail to clear the organism, and, paradoxically, may participate in and perpetuate an immune and inflammatory response leading to pulmonary injury.25–27 Hence, a high antibody titre and formation of immune complexes are poor prognostic features.28 Our studies provide no evidence that serum IgG B. cepacia antibodies have the same significance for lung injury in those colonized by B. cepacia. These antibodies do not appear to achieve clearance of B. cepacia from the lung, as evidenced by continuing colonization in the face of increasing antibody titres. However, our study does not support the hypothesis that high antibody titres are associated with poor prognosis. This conclusion relates only to the particular antibody utilized in the study. It is possible that antibodies to different antigenic components of B. cepacia, including flagellar and Oside chain antigen, may have different pathophysiological significance. While it is clear that production of IgG antibodies against B. cepacia LPS does not clear the organism from the lungs, it is not clear whether the presence of this antibody in the serum of patients colonized with P. aeruginosa may afford future protection against pulmonary colonization with B. cepacia. This study did not indicate any difference in B. cepacia LPS antibody titres between P. aeruginosa colonized patients who subsequently became colonized with B. cepacia, and therefore suggests no protective interaction. The division of B. cepacia patients into those chronically colonized and those who grew the organism intermittently
169
was based upon the definition of colonization without taking into account differences in clinical behaviour. Greater B. cepacia antibody titres in the colonized group indicates a heightened humoral immune response to the organism compared to the intermittent growers, though antibody levels in the intermittent growers of B. cepacia were greater than in those colonized with P. aeruginosa, P\0.05. An alternative explanation for our findings is that the antibody response to B. cepacia core antigen is related to the ‘burden’ or duration of infection, and may be only partially related to the pathophysiology of lung injury. In this respect, these findings are similar to those relating to antibody patterns in chronic pulmonary infection with P. aeruginosa in CF. The degree of overlap of antibody titres demonstrated here does not allow antibody levels to be used in a clinical context to define infection status in patients with CF.
References 01. Isles A, Macluskey, I, Corey M. Pseudomonas cepacia infection in cystic fibrosis: an emerging problem. J Paediatr 1984; 104: 206–210. 02. Thomassen MJ, Demko CA, Klinger JD, Stern RC. Pseudomonas cepacia: colonisation among patients with cystic fibrosis. Am Rev Resp Dis 1985; 131: 791–796. 03. Simmonds EJ, Conway SP, Ghonheim ATM, Littlewood JM. Pseudomonas cepacia: a new pathogen in patients with cystic fibrosis referred to a large centre in the United Kingdom. Arch Dis Child 1990; 65: 874–877. 04. Gessner AR, Mortensen JE. Pathogenic factors of Pseudomonas cepacia isolates from patients with Cystic Fibrosis. J Med Microbiol 1990; 33: 115–120. 05. Govan JRW, Hughes JE, Vandamme P. Burkholderia cepacia: medical, taxonomic and ecological issues. J Med Microbiol 1996; 45: 385–407. 06. Nelson JW, Butler SL, Kreig D, Govan JRW. Virulence factors of Burkholderia cepacia. FEMS Immunol Med Microbiol 1994; 8: 89–97. 07. Stover G, Drake DR, Montie TC. Virulence of different Pseudomonas species in a burned mouse model: tissue colonisation by Pseudomonas cepacia. Infect Immun 1983; 41: 1099–1144. 08. Suter S, Schaad UB, Roux-Lombard P, Girardin E, Grau G, Dayer JM. Relation between TNF-a and granulocyte elastase-a1-proteinase inhibitor complexes in plasma of patients with cystic fibrosis. Am Rev Resp Dis 1989; 140: 1640–1644. 09. Suter S, Schaad Uh, Roux L, Nydegger U, Waldvogel F. Granulocyte neutral proteases and pseudomonal elastase as possible causes of airway damage in patients with cystic fibrosis. J Infect Dis 1984; 149: 523–531. 10. Meyer K, Lewandowski J, Zimmerman J, Nunley D, Calhoun W, Dopico G. Human neutrophil elastase a-1-antiprotease complex in cystic fibrosis. Am Rev Resp Dis 1991; 144: 580–585. 11. Hughes JE, Stewart J, Barclay GR, Govan JRW. Priming of neutrophil burst activity by lipopolysaccharide of Burkholderia cepacia. Infect Immun 1997; 65: 4281–4287. 12. Hutchinson ML, Poxton IR, Govan JRW. Burkholderia cepacia produces a haemolysin that is capable of inducing apoptosis and degranulation of mammalian phagocytes. Infect Immun 1998; 66: 2033–2039. 13. Zughaier SM, Ryley HC, Jackson SK. A melanin pigment purified from an epidemic strain of Burkholderia cepacia attenuates monocyte respiratory burst activity by scavenging superoxide anion. Infect Immun 1999; 67: 908–913.
170
J. Hendry et al.
14. Zughaier SM, Ryley HC, Jackson SK. Lipopolysaccharide (LPS) from Burkholderia cepacia is more active than LPS from Pseudomonas aeruginosa and Stenotrophomonas maltophilia in stimulating tumour necrosis factor alpha from human monocytes. Infect Immun 1999; 67: 1505–1507. 15. Aronoff SC, Quiner FJ, Stern RC. Longitudinal serum IgG response to Pseudomonas cepacia surface antigens in cystic fibrosis. Paediatr Pulmonol 1991; 1: 289–293. 16. Aronoff SC, Stern RC. Serum IgG antibody to the outer membrane antigens of Pseudomonas cepacia and Pseudomonas aeruginosa in cystic fibrosis. J Infect Dis 1988; 157: 934–940. 17. Lacy D, Smith A, Stableforth D, Smith G, Weller P, Brown M. Serum IgG response to B. cepacia outer membrane antigens in cystic fibrosis: assessment of cross-reactivity with P. aeruginosa. FEMS Immunology and Medical Microbiology 1995; 10: 253–262. 18. Nelson JW, Butler SL, Brown PH, Greening AP, Govan JRW. Serum IgG and sputum IgA antibody to core lipopolysaccharide antigen from Pseudomonas cepacia in patients with cystic fibrosis. J Med Microbiol 1993; 39: 39–47. 19. Nelson JW, Govan JRW, Barclay GR. Pseudomonas aeruginosa flagellar antibodies in serum, saliva and sputum from patients with cystic fibrosis. Serodiag Immunother Inf Dis 1990; 4: 351–361. 20. Brett MM, Ghonheim ATM, Littlewood JM. Serum IgG antibodies in patients with cystic fibrosis with early P. aeruginosa infection. Arch Dis Child 1987; 62: 357–361. 21. Doring G, Hoiby N. Longitudinal study of immune response to Pseudomonas aeruginosa antigens in cystic fibrosis. Infect Immun 1983; 42: 197–201.
22. Nelson JW, Butler SL, Brown PH, Greening AP, Govan JRW. Serum IgG and sputum IgA antibody to core lipopolysaccharide antigen from Pseudomonas cepacia in patients with cystic fibrosis. J Med Microbiol 1993; 39: 39–47. 23. Hollsing AE, Granstrom M, Vasil ML et al. Prospective study of serum antibodies to Pseudomonas aeruginosa exoproteins in cystic fibrosis. J Clin Microbiol 1987; 25: 1830–1836. 24. Nelson JW, Barclay R, Govan JRW. Diagnosis of chronic Pseudomonas aeruginosa infection in cystic fibrosis by enzyme linked immunosorbent assay for anti pseudomonas LPS antibodies. Serodiag Immunother Infect Dis 1990; 4: 9–16. 25. Moss RB. Hypergammaglobulinaemia in cystic fibrosis. Chest 1987; 91: 522–526. 26. Wheeler WB, Williams M, Matthews WJ, Colten HR. Progression of cystic fibrosis lung disease as a function of serum immunoglobulin levels: a five year longitudinal study. J Paediatr 1984; 104: 695– 699. 27. Tosi M, Zakem-Cloud H, Demko C et al. Cross-sectional and longitudinal studies of naturally occuring antibodies to Pseudomonas aeruginosa in cystic fibrosis. J Inf Dis 1995; 172: 453–461. 28. Manthei U, Taussig LM, Beckerman RC, Strunk RC. Circulating immune complexes in cystic fibrosis. Am Rev Resp Dis 1982; 126: 253–257.