Thorax 1994;49:803-807

Effect of antibiotic treatment on inflammatorymarkers and lung function in cystic fibrosispatients with Pseudomonas cepacia

D Peckham, S Crouch, H Humphreys, B Lobo, A Tse, A J Knox

AbstractBackground - The acquisition of Pseudo-monas cepacia in patients with cysticfibrosis is associated with increasingdeterioration in lung function and morefrequent hospital admissions. Pseudomo-nas cepacia is usually resistant to severalantibiotics in vitro, but the response ofpatients colonised with the organism hasnot been extensively studied in vivo.Methods - A three month prospectivestudy was performed to investigate theresponse of 14 Ps cepacia positive patientsand 10 Ps cepacia negative patients to atwo week course of intravenous antibi-otics. All those who were Ps cepacia nega-tive and six of the 14 Ps cepacia positivepatients had Ps aeruginosa in their spu-tum which was sensitive to the prescribedtherapy. The inflammatory markers C-reactive protein, white blood cell count,serum lactoferrin, neutrophil elastase/ae-antitrypsin complex, and tumour necrosisfactor alpha were measured at the startand end of each antibiotic course.Results - The median (range) % improve-ment in baseline FEV, and FVC followingtreatment in the group as a whole was15-2% (-23-5% to 156-3%) and 23-9%(-36-8% to 232-7%) respectively. Therewas no statistical difference in improve-ment in lung function, body weight, orinflammatory markers between indi-viduals who were Ps cepacia positive andthose who were Ps cepacia negative.Conclusions - Patients who are Ps cepaciapositive appear to respond as well tointravenous antibiotics as those who arePs cepacia negative, despite having lowerlung function and a bacterium in theirsputum which is resistant in vitro to theantibiotics used.

(Thorax 1994;49:803-807)

In the UK the incidence and prevalence ofPseudomonas cepacia among patients with cysticfibrosis has significantly increased over the pastfew years.'-3 This can partly be explained byimprovement in microbiological isolation tech-niques and by an increase in both social andhospital contacts among patients with cysticfibrosis.2 Despite the increased evidence forpatient to patient transmission, it remainsunclear whether Ps cepacia is a cause of, ormarker of, disease severity.

In a small proportion of patients with Ps

cepacia rapid and unexpected lung deteriorationdevelops despite intensive antibiotic treatment.The isolation of Ps cepacia from blood culturesaccompanied by clinical evidence of systemicsepsis provides evidence of a possible patho-genic role for this organism. 145 In mostpatients, however, a gradual decline in lungfunction occurs following the isolation of theorganism.6

Pseudomonas cepacia is resistant to mostantipseudomonal antibiotics in vitro.7 Our clin-ical impression is nevertheless that patientsrespond clinically to these antibiotics. Markerssuch as serum levels of neutrophil elastase/ot,-antitrypsin complex, lactoferrin, tumour necro-sis factor alpha (TNFca), C-reactive protein,and white blood cell count have been shownpreviously to reflect the inflammatory processin patients with cystic fibrosis and are reducedfollowing effective antipseudomonal treatmentin patients colonised with Ps aeruginosa.8'0 It isnot known, however, whether similar changesare seen after treatment in patients colonisedwith Ps cepacia. We have therefore prospect-ively studied the effectiveness of intravenousantibiotic therapy in a group of adult patientswith cystic fibrosis, with and without Ps cepa-cia, measuring treatment response as change ininflammatory markers, lung function, and bodyweight.

MethodsSTUDY DESIGNThis was a prospective study which included allpatients who required two weeks of intravenousantibiotics for infective exacerbations of cysticfibrosis over a three month period at Notting-ham City Hospital. Patients were assessed bothbefore and at the end of the two week course ofantibiotics with measurements of body weight,lung functions, and venous blood samples formeasurements of inflammatory markers.

PATIENTSTwenty four adult patients were studied, 10 ofwhom were colonised with Ps aeruginosa with-out Ps cepacia (Ps cepacia negative) and 14 ofwhom were colonised with Ps cepacia, with orwithout Ps aeruginosa (Ps cepacia positive).Clinical details are outlined in table 1. Allpatients had been chronically infected with Psaeruginosa for more than two years. Of thepatients who were Ps cepacia positive, thisbacterium had been isolated from one patientthree weeks before antibiotic therapy while theremaining 13 had repeatedly been positive for

Division ofRespiratory Medicine,Medical ResearchCentre, City Hospital,Nottingham NG5 IPBD PeckhamS CrouchB LoboA TseA J Knox

Department ofMicrobiology, QueensMedical Centre,University Hospital,Nottingham NG7 2UHH Humphreys

Reprint requests to:Dr D Peckham.Received 7 February 1994Returned to authors5 April 1994Revised version received20 April 1994Accepted for publication26 April 1994

803

on 11 Septem

ber 2018 by guest. Protected by copyright.

http://thorax.bmj.com

/T

horax: first published as 10.1136/thx.49.8.803 on 1 August 1994. D

ownloaded from

Peckham, Crouch, Humphreys, Lobo, Tse, Knox

Table I Clinical details of 24 adult patients with cystic fibrosis

Ps cepacia positive Ps cepacia negative p(n= 14) (n= 10)

Mean age (years) 22 8 22-2M:F 10:4 6:4Schwachman score 45 (25-70) 65 (30-80) < 0-05Chrispin-Norman score 20 (16-32) 20 (9-32) > 0 05Post-treatment FEV, (l/s) 1 17 (0-58-221) 1 66 (0 714-93) 0 05% predicted FEV 31 (14-64) 48 (17-88) > 0-05Post-treatment FkC (1) 1 75 (0-76-35) 2 73 (1 66-5-64) <0-001% predicted FVC 46 (16-70) 65 (35-79) < 0 05

Values are median (range).

Ps cepacia for more than six months. Twopatients who were Ps cepacia positive and onewho was Ps cepacia negative were on low doseoral prednisolone prior to antibiotic therapy(10-20 mg/day). None of the patients werestarted on steroids over the three month studyperiod. Four patients in both groups receivedtheir antibiotics at home while being reviewedweekly on the ward and the remainder weretreated as inpatients. Pathogens isolated fromsputum before the start of intravenous anti-biotic therapy are summarised in table 2.

LUNG FUNCTIONForced expiratory volume in one second (FEVy)and forced vital capacity (FVC) were measuredas the highest of three blows on a Vitalograph aspirometer (Buckingham, UK).

MICROBIOLOGYNeat sputum and a 1:10 000 dilution of sputumfrom patients with cystic fibrosis was routinelycultured for Haemophilus influenzae, Staphylo-coccus aureus, and Ps aeruginosa following diges-tion with N-acetyl cysteine using cefsulodinchocolate agar and MacConkey agar. Investiga-tion for the presence of mycobacteria and atypi-cal respiratory pathogens was carried out whenindicated. Sputum samples were also inocu-lated on to Ps cepacia selective agar medium(MAST, UK) incorporating ticarcillin(100mg/l) and polymyxin B (30000 units/l).Agar plates were incubated for 40 hours at 37°Cand Ps cepacia was identified by colonialappearance, oxidase reaction, biochemical reac-tion, and resistance to polymyxin B. Antimicro-bial susceptibility testing to gentamicin, tobra-mycin, ceftazidime, azlocillin, ciprofloxacin,aztreonam, imipenem, and polymyxin B wascarried out using the disc diffusion method."

ANTIBIOTIC TREATMENTAll patients who were Ps cepacia negative

Table 2 Organisms isolated from sputum of patientsbefore treatment with intravenous antibiotics

Frequency

Ps cepacia negativePs aeruginosa alone 8Staph aureus and Ps aeruginosa 1H influenzae and Ps aeruginosa 1

Ps cepacia positivePs cepacia alone 6Ps cepacia and Ps aeruginosa 6Ps cepacia, Staph aureus and H influenzae 1H influenzae and Ps cepacia 1

received an aminoglycoside (gentamicin ortobramycin) with 5 g three times daily azlocillin(one patient), or 2 g three times daily ceftazi-dime (seven patients), 2 g three times dailyaztreonam (one patient) or 1 g three times dailyimipenem alone (one patient). Patients whowere Ps cepacia positive received a combinationof an aminoglycoside with either ceftazidime(eight patients), azlocillin (five patients), oraztreonam (one patient) at the identical doses topatients who were Ps cepacia negative. Anti-biotic regimens were selected according tosensitivity results of both Ps aeruginosa and Pscepacia, and in patients who were Ps cepaciapositive combination therapy was used. Thechoice of antibiotics was arbitrary when a mul-tiresistant strain of Ps cepacia was isolated in theabsence of Ps aeruginosa. At the start of treat-ment all 10 isolates of Ps aeruginosa amongpatients who were Ps cepacia negative and fiveisolates of Ps aeruginosa from patients who werePs cepacia positive were fully sensitive to theantibiotics used. One patient who was Ps cepa-cia positive grew a Ps aeruginosa isolate whichproved to be sensitive to tobramycin alone. Ofthe 14 patients who were Ps cepacia positive 12had a multiresistant strain of Ps cepacia in theirsputum while the isolates from two patientswere sensitive to ceftazidime alone. Twopatients in the Ps cepacia positive group and onein the Ps cepacia negative group received eitherhigh dose oral amoxicillin (3 g twice daily) orintravenous cefuroxime (1 5 g three times daily)to treat additional H influenzae (table 2), whileone patient in each group was treated with oralflucloxacillin (1 g four times daily) for addi-tional Staphylococcus aureus infection. Serumaminoglycoside levels were measured aroundthe fourth dose and again on the third or fourthday thereafter, with dose adjustment as appro-priate to maintain a serum peak concentrationof 7-10 mg/l. Five patients who were Ps cepacianegative and five who were Ps cepacia positivewere on long term nebulised antibiotic therapywhich was discontinued during the studyperiod. Of the patients who were Ps cepacianegative two were on colomycin (1 megaunittwice daily), two on gentamicin (80 mg twicedaily) and one on tobramycin (80 mg twicedaily), while amongst patients who were Pscepacia positive four were on colomycin (1megaunit twice daily) and one was on gentami-cin (80 mg twice daily).

INFLAMMATORY MARKERSFull blood count with differential counts wasmeasured by conventional automated analysis,C-reactive protein by a nephelometry method,'0and other inflammatory markers by enzymelinked immunosorbent assay (ELISA) follow-ing serum storage at - 70°C. The lactoferrinELISA used'2 had a lower detection limit of0-005 nmol/l. Immunoreactive TNF was mea-sured according to a previously describedmethod'3 with a lower detection limit of6*25 pg/ml. Elastase/a,-antitrypsin was mea-sured using a modification of a previously de-scribed method'4 where the streptavidinhorseradish peroxidase step was replaced with

804

on 11 Septem

ber 2018 by guest. Protected by copyright.

http://thorax.bmj.com

/T

horax: first published as 10.1136/thx.49.8.803 on 1 August 1994. D

ownloaded from

Effect of antibiotic treatment on inflammatory markers and lung function in cystic fibrosis patients with Ps cepacia

Table 3 Median (range) results for inflammatory markers before and after treatment in the combined group of24 adult patients with cystic fibrosis

Inflammatory markers Before treatment After treatment p

White cell count (x 109/l) 10-9 (24-21 5) 8-7 (20-18 3) 0 005C-reactive protein (mg/I) 29 (< 11-169) <11 mg/l (< 11-130) 0 001Lactoferrin (nmol/1) 6 97 (3 87-8-99) 4 87 (3 29-787) 0.02Tumour necrosis factor (pg/ml) 42 25 (0-1055) 27 2 (0-604 5) > 0-05Elastase/a,-antitrypsin complex (ng/ml) 728 (277-1541) 407 5 (54-842) <0 001

an alkaline phosphatase conjugate to strept-avidin. An AMPAK amplification kit (Dako)was used to increase the sensitivity of the assayallowing detection of the complex to lOng/mlin the serum.

DATA ANALYSISBecause the data were not normally distributedthe results were expressed as medians andranges. Changes in weight, lung function, andinflammatory markers were analysed in patientswho were Ps cepacia positive, those who werePs cepacia negative, and both groups combinedusing Wilcoxon and Mann-Whitney tests forpaired and unpaired data respectively (Micro-soft Corporation, Redmond, USA). Subgroupanalysis comparing the same variables in the Pscepacia positive group (six patients Ps aerugi-nosa positive, six patients Ps aeruginosa nega-tive) was also carried out. A p value < 0 05 wasregarded as statistically significant.

ResultsPatients who were Ps cepacia positive had signi-ficantly lower median FEV1, FVC, andSchwachman scores after treatment thanpatients who were Ps cepacia negative, althoughChrispin-Norman scores were similar (table 1).

SPIROMETRYWhen patients who were Ps cepacia positive andPs cepacia negative were combined lung func-tion improved significantly following antibiotictreatment from a median (range) % predictedFEV1 and FVC before therapy of 28-5% (11-67%) and 38% (9-480%) respectively to amedian (range) FEV1 of 34 5% (14-88%,p < 005) and FVC of 53% (16-79%, p < 005).The median % improvement of FEV, and FVCfollowing antibiotics was 15-23% (- 23 46% to156-25%) and 23 99% (- 36 8% to 232-7%)respectively.There was no difference in % improvement

of FEV1 and FVC before and after treatmentwhen comparing patients who were Ps cepaciapositive and negative. The median (range) %improvement of FEV1 and FVC from baselinein patients who were Ps cepacia positive was19-49% (-23-46% to 156 25%, p>0 05) and23 35% (-36 8% to 232-65%, p<0 05) re-

spectively, and 8 075% (-7 73% to 99-6%,p>0 05) and 23-99% (-3 9% to 145 9%,p > 0 05) in patients who were Ps cepacia nega-tive.There was no difference between the % im-

provement in both FEV1 and FVC among thesix patients who were Ps aeruginosa and Pscepacia positive, and the six patients who werePs cepacia positive but Ps aeruginosa negative.

WEIGHTThe median (range) weight of all patients was53-2 (39-1-72-5) kg before treatment and 54-2(40 6-76 6) kg (p=0O001) after treatment. Themedian improvement in weight was 0 5 (- 12to 4 7) kg.The median weight of patients who were Ps

cepacia positive was 53-2 (39-1-66-3) kg beforeand 54 1 (40 6-68 8) kg after treatment whereasin patients who were Ps cepacia negative thecorresponding results were 53 4 (44 6-72 5) kgand 55 3 (45-76-6) kg respectively. The dif-ference in change in weight between patientswho were Ps cepacia positive and negative wasnot significant. There was no difference in thechange in weight during treatment when the sixpatients with Ps aeruginosa and Ps cepacia werecompared with the six patients with Ps cepaciaalone. Seven of the patients who were Ps cepaciapositive were on long term nasogastric feedingcompared with only two of the patients whowere Ps cepacia negative.

INFLAMMATORY MARKERSThe results before and after treatment for thecombined group are outlined in table 3. Therewas a significant fall in white blood cell countand serum levels of C-reactive protein, lacto-ferrin and acx-antitrypsin following treatmentfor the combined group. No differences weredetected in the parameters studied betweenpatients who were Ps cepacia positive and thosewho were Ps cepacia negative or between thetwo Ps cepacia subgroups (six Ps aeruginosapositive, six Ps aeruginosa negative). No signi-ficant difference was seen in the pretreatment,post-treatment, or in the change in any of themeasured parameters. The median changes fol-lowing treatment in the two groups are outlinedin table 4.

Table 4 Median (range) change in inflammatory markers following intravenous antibiotic therapy in 14 patients whowere Ps cepacia positive and 10 patients who were Ps cepacia negative

Change in inflammatory Ps cepacia positive Ps cepacia negative pmarkers following treatment

White cell count (x 10'/l) -1.1 (2-9 to -92) -1-6 (1-2 to -8-0) >0-5C-reactive protein (mg/I) -32 (5 to -144) -14 (21 to -169) >0.5Lactoferrin (nmol/1) -003 (1-93 to -0-56) - 0.12 (0-80 to - 0-56) >0.5Tumour necrosis factor(pgpml) - 3875 (53 to - 235) 022 (196- to - 4505) >0 5Elastase/a,-antitrypsin complex (ng/ml) - 350 (168 to - 752) -407 (100 to -836) > 05

The p value shown is for the difference in the change in each variable between patients who were Ps cepacia positive and those whowere Ps cepacia negative.

805

on 11 Septem

ber 2018 by guest. Protected by copyright.

http://thorax.bmj.com

/T

horax: first published as 10.1136/thx.49.8.803 on 1 August 1994. D

ownloaded from

Peckham, Crouch, Humphreys, Lobo, Tse, Knox

When serum lactoferrin levels are correctedfor the number of circulating neutrophils themedian (range) for the 24 patients before treat-ment was 0-885 (0 503-1-479) nmol/106 neutro-phils and 0 947 (0 368-2636) nmol/106 neutro-phils following treatment (p > 0-05). Therewere no significant differences between cor-rected values of serum lactoferrin and TNFlevels before and after treatment when patientswho were Ps cepacia positive and negative werecompared (p > 0 05).

DiscussionPseudomonas cepacia has been isolated withincreasing frequency in specialist centres withinthe UK.' 2 Isolation of Ps cepacia from patientswith cystic fibrosis is associated with poor lungfunction, increasing age, recent hospitalisation,and close hospital or social contact with otherpatients who are Ps cepacia positive and sib-lings.'5 Despite some evidence to suggest thatcolonisation with Ps cepacia heralds a poorerprognosis, it is unclear whether this is becausethe organism is pathogenic or because it is amarker for increased disease severity due toother factors.'5 Pseudomonas cepacia is non-pathogenic in healthy individuals and in animalmodels it has been found to be relatively aviru-lent compared with Ps aeruginosa. '6'7 It is pos-sible that Ps cepacia acts synergistically withother bacteria to exacerbate chest disease incystic fibrosis.

In this study we have investigated the effectof intravenous antipseudomonal antibiotictreatment in patients with Ps cepacia infectionand have compared this with patients colonisedby Ps aeruginosa. Several parameters were usedto assess response including lung function andvarious inflammatory markers. In our group of24 patients studied prospectively we found thatlung function and Schwachman scores werelower in patients colonised with Ps cepacia thanin those with Ps aeruginosa alone. This is con-sistent with the results of other studies andsuggests an association between colonisationwith Ps cepacia and poor clinical status.418 De-spite the fact that patients who are Ps cepaciapositive had worse lung function and Schwach-man scores, antibiotic treatment was equallyeffective in improving lung function and weightin both groups. Both these parametersimproved significantly over the two weeks oftreatment in both groups of patients and wassimilar to that reported in previous studies ofpatients colonised with Ps aeruginosa.8'920 Thefact that weight improved equally in patientswho were Ps cepacia positive despite their gen-erally poorer clinical status may partially reflectthe fact that more of them were on nasogastricnutritional supplementation.

Several inflammatory markers weremeasured including white blood cell count andserum levels of C-reactive protein, lactoferrin,neutrophil elastase/acl-antitrypsin complex, andTNF. These have all been shown previously toreflect the inflammatory process in patientswith cystic fibrosis. Previous studies of patientscolonised with Ps aeruginosa have shownthat values of these markers fall following

treatment with intravenous antipseudomonalantibiotics."'°The inflammatory markers elastase/oxi-anti-

trypsin, lactoferrin, and C-reactive protein fellin response to intravenous antibiotic therapy inboth patients who were Ps cepacia positiveand those who were Ps cepacia negative. Thechanges seen were similar to those found pre-viously in studies of patients colonised with Psaeruginosa.810 No significant change in TNFlevels occurred following antibiotic therapy ineither group, but many patients had undetec-table levels of TNF both before and aftertreatment. Previous studies have expressed lac-toferrin levels as absolute values. When we didthe same we found a significant reduction inlactoferrin levels with antibiotic treatment. Thesole source of lactoferrin is the neutrophil,however, and neutrophil values also fell.2122When the lactoferrin level was adjusted for theneutrophil count no significant change wasseen. This suggests that the fall in lactoferrinlevels is largely due to the fall in neutrophilcount. As with the lung function results, therewas no significant difference in improvement inany of the inflammatory markers betweenpatients who were Ps cepacia positive and thosewho were Ps cepacia negative.The good clinical response to intravenous

antibiotic therapy among patients who were Pscepacia positive is surprising in the light of thein vitro sensitivity pattern seen with Ps cepacia.The organism is resistant to most commonlyused antipseudomonal agents,7 but com-binations of antibiotics - for example, amino-glycosides such as gentamicin and a 1-lactamagent such as azlocillin - may, however, inhibitthe growth of Ps cepacia synergistically. In vitrocombinations of other antibiotics have beenshown to be synergistic against Ps cepacia.2325Synergy between three antibiotics (rifampicin,imipenem, and ciprofloxacin) has previouslybeen demonstrated and therefore two or moreantimicrobial agents may be needed.25 Manypatients with cystic fibrosis who are colonisedwith Ps cepacia also carry Ps aeruginosa andexacerbation may be caused by Ps aeruginosawith or without H influenzae or Staph aureus.Our clinical experience would seem to indicatethat combination chemotherapy results in invivo activity against Ps cepacia despite resist-ance in vitro. Nottingham isolates of Ps cepaciaare resistant to the aminoglycosides, azlocillin,ciprofloxacin, imipenem, aztreonam, and poly-myxin B, but some are moderately sensitive toceftazidime and most adult patients with cysticfibrosis show the same strain (unpublished ob-servation).We have considered several explanations for

the clinical improvement of patients who werePs cepacia positive despite in vitro antimicrobialresistance. Firstly, it is possible that in vitroantibiotic sensitivities do not reflect the in vivosituation within the lung due to the milieu ofthe inflammatory response. Alternatively the invivo response may be due to synergism betweenantibiotics. Current methodology used in anti-biotic susceptibility testing of Ps cepacia may beless appropriate for this organism which isslower growing than Ps aeruginosa and grows

806

on 11 Septem

ber 2018 by guest. Protected by copyright.

http://thorax.bmj.com

/T

horax: first published as 10.1136/thx.49.8.803 on 1 August 1994. D

ownloaded from

Effects of antibiotic treatment on inflammatory markers and lung function in cystic fibrosis patients with Ps cepacia

preferentially at 30°C. It is also possible that aheavy growth of Ps cepacia may inhibit therecognition of other organisms such as Staphaureus and Haemophilus. Whilst this is a possib-ility it would not explain the results of treat-ment directed towards Ps aeruginosa in ourstudy. An alternative explanation is that theresponse in both patient groups is due solely toeffective physiotherapy which is causing im-provement by enhancing sputum expectorationand reducing the inflammatory stimulus, thusreducing levels of inflammatory markers. Thislatter possibility would seem unlikely as pre-vious studies have shown that intravenous anti-biotics alone are more effective than physiother-apy alone at reducing Ps aeruginosa counts insputum.20 Antibiotics are also known to haveother effects which may modify the response toinfection including the release of endotoxin andthe inhibition of the cytokine cascade.26 Thepossible immunomodulatory effect combinedwith some in vivo antibacterial activity, physio-therapy, and nutritional support may explainthe clinical improvement. Consequently, bene-ficial effects might be explained despite persist-ence of the organism. Lastly, we consideredwhether the response to treatment in patientscolonised with Ps cepacia may reflect treatmentof Ps aeruginosa rather than Ps cepacia. For thisreason we have compared the two subgroups ofpatients who were Ps cepacia positive (with andwithout Ps aeruginosa) and have found no dif-ferences in either lung function or inflam-matory markers between patients colonisedwith Ps aeruginosa and Ps cepacia and thosecolonised with Ps cepacia alone. This suggeststhat treatment was as effective when only Pscepacia was present. Alternatively, the clinicalresponse to antibiotic treatment among patientswho appeared to be colonised by Ps cepaciaalone may simply reflect the treatment ofunderlying Ps aeruginosa as the inability toisolate this organism from the sputum does notexclude its presence within the lower respira-tory tract.The fact that we found that patients who

were Ps cepacia positive responded well toantibiotics suggests that the reason for thegreater decline in their lung function insome studies may be the result of lung damageoccurring between antibiotic courses.6 It isinteresting that our patients who were Ps cepaciapositive required more courses of intravenousantibiotics over the previous 12 months thanthose patients who were Ps cepacia negative.We conclude, therefore, that antibiotic treat-

ment of patients who are Ps cepacia positive isoften effective in vivo despite the multiresistantnature of this organism in vitro. Further work isrequired to determine the relation between invitro sensitivity results and the response in vivoand the pattern of inflammatory response over aprolonged period in patients who are Ps cepaciapositive. From a practical point of view anti-

biotics should not be withheld from patientsbecause of in vitro resistant patterns.

1 Simmonds EJ, Conway SP, Ghoneim ATM, Ross H, Little-wood JM. Pseudomonas cepacia: a new pathogen in patientswith cystic fibrosis referred to a large centre in the UnitedKingdom. Arch Dis Child 1990;65:874-7.

2 Govan JRW, Brown PH, Maddison J, Doherty CJ, NelsonJW, Dodd M, et al. Evidence for transmission of Pseudo-monas cepacia by social contact in cystic fibrosis. Lancet1993;342: 15-9.

3 Editorial. Pseudomonas cepacia - more than a harmless com-mensal? Lancet 1992;339:1385-6.

4 Lewin LO, Byard PJ, Davis PB. Effect of Pseudomonascepacia colonization on survival and pulmonary function ofcystic fibrosis patients. J Clin Epidemiol 1990;43:125-31.

5 Tablan OC. Nosocomially acquired Pseudomonas cepaciainfection in patients with cystic fibrosis. Infect ControlHosp Epidemiol 1993;14:124-6.

6 Muhdi KM, O'Hickey S, Smith G, Stableforth DE. Rate ofdecline of lung function in patients with cystic fibrosiscolonised by Pseudomonas cepacia. Thorax 1993;48:1088.

7 Bhakta DR, Leader I, Jacobson R, Robinson-Dunn B,Honicky RE, Kumar A. Antibacterial properties of investi-gational, new, and commonly used antibiotics against isol-ates of Pseudomonas cepacia in Michigan. Chemotherapy1992;38:319-23.

8 Rayner RJ, Wiseman MS, Cordon SM, Norman D, HillerEJ, Shale DJ. Inflammatory markers in cystic fibrosis.Respir Med 1991;85:139-45.

9 Norman D, Elborn JS, Cordon SM, Rayner RJ, WisemanMS, Shale DJ. Plasma tumour necrosis factor alpha incystic fibrosis. Thorax 1991;46:91-5.

10 Valletta EA, Rigo A, Bonazzi L, Zanolla L, Mastella G.Mbdification of some markers of inflammation duringtreatment for acute respiratory exacerbation in cystic fibro-sis. Acta Paediatr 1992;81:227-30.

11 Holt A, Brown D. Antimicrobial susceptibility. In: HawkeyPM, Lewis DA, eds. Medical bacteriology. A practicalapproach. Oxford: Oxford University Press, 1989:167-94.

12 Crouch SPM, Fletcher J. Effect of ingested pentoxifylline onneutrophil superoxide anion production. Infect Immun1992;60:4504-9.

13 Meager A, Partis S, Leung H, Piel E, Mahon B. Preparationand characterisation of monoclonal antibodies directedagainst antigenic determinants of recombinant humantumour necrosis factor (rTNF). Hybridoma 1987;6:305-1 1.

14 Rocker GM, Wiseman MS, Pearson D, Shale DJ. Diagnosticcriteria for adult respiratory distress syndrome: time forreappraisal. Lancet 1989;i: 120-3.

15 Gilligan PH. Microbiology of airway disease in patients withcystic fibrosis. Clin Microbiol Rev 1991;4:35-51.

16 Goldmann DA, Klinger JD. Pseudomonas cepacia: biology,mechanisms of virulence, epidemiology. J Pediatr 1986;108:806-12.

17 Starke JR, Edweards MS, Langston C, Baker CV. A mousemodel of chronic pulmonary infection with Pseudomonasaeruginosa and Pseudomonas cepacia. Pediatr Res1987;22:698-702.

18 Tablan OC, Martone WJ, Doershuk CF, Stern RC, Tho-massen MJ, Klinger JD, et al. Colonization of the respira-tory tract with Pseudomonas cepacia in cystic fibrosis. Riskfactors and outcomes. Chest 1987;91:527-32.

19 Hodson ME, Roberts CM, Butland RJA, Smith MJ, BattenJC. Oral ciprofloxacin compared with conventional intra-venous treatment for Pseudomonas aeruginosa infection inadults with cystic fibrosis. Lancet 1987;i:235-7.

20 Regelmann WE, Elliott GR, Warwick WJ, Clawson CC.Reduction of sputum Pseudomonas aeruginosa density byantibiotics improves lung function in cystic fibrosis morethan do bronchodilators and chest physiotherapy alone.Am Rev Respir Dis 1990;141:914-21.

21 Baynes RD, Bezwoda WR, Khan Q, Mansoor N. Relation-ship of plasma lactoferrin content to neutrophil regen-eration and bone marrow infusion. Scand J Haematol1986;36:79-84.

22 Boxer LA, Smolen JE. Neutrophil granule constituents andtheir release in health and disease. Hematol Oncol ClinNorth Am 1988;2:101-20.

23 Kumar A, Wofford-McQueen R, Gordon RC. In vitroactivity of multiple antimicrobial combinations againstPseudomonas cepacia isolates. Chemotherapy 1989;35:246-53.

24 Chin NX, Neu HC. Synergy of new C-3 substituted cepha-losporins and tobramycin against Pseudomonas aeruginosaand Pseudomonas cepacia. Diagn Microbiol Infect Dis1989;12:343-9.

25 Kumar A, Wofford-McQueen R, Gordon RC. Ciprofloxa-cin, imipenem and rifampicin: in-vitro synergy of two andthree drug combinations against Pseudomonas cepacia. JAntimicrob Chemother 1989;23:831-5.

26 Simon DM, Koenig G, Trenhome GM. Differences inrelease of tumour necrosis factor from TPH-1 cells stimu-lated by filtration of antibiotic-killed Escherichia coli. JInfect Dis 1991;164:800-2.

807

on 11 Septem

ber 2018 by guest. Protected by copyright.

http://thorax.bmj.com

/T

horax: first published as 10.1136/thx.49.8.803 on 1 August 1994. D

ownloaded from