Controlled Trial of Inhaled Budesonide in Patients with ... - ATS Journals

Controlled Trial of Inhaled Budesonide in Patients with Cystic Fibrosis and Chronic Bronchopulmonary Pseudomonas aeruginosa Infection HANS BISGAARD, SVEND STENVANG PEDERSEN, KIM GJERUM NIELSEN, MARIANNE SKOV, EVA MOSFELDT LAURSEN, GITTE KRONBORG, CLAUS MICHAEL REIMERT, NIELS HØIBY, and CHRISTIAN KOCH Departments of Pediatrics, Infectious Diseases, and Clinical Microbiology and Laboratory of Medical Allergology, National University Hospital, Rigshospitalet, Copenhagen, Denmark

The efficacy and safety of anti-inflammatory treatment with inhaled glucocorticosteroids in patients with cystic fibrosis (CF) and complicating chronic Pseudomonas aeruginosa (P.a.) lung infection was studied in a placebo-controlled, parallel, double-blind single center trial. Active treatment consisted of budesonide dry powder, 800 mg twice daily, delivered from a Turbuhaler®. The study period covered two successive 3-mo intervals between elective courses of intravenous anti-Pseudomonas antibiotics. Fifty-five patients entered the study, with a mean age of 20 yr and a mean FEV1 of 63% of predicted. Analysis of all patients entered, irrespective of trial adherence (“intention to treat”), showed a decrease in FEV1 in the first period of 20.032 L in patients on budesonide versus 20.187 L in patients on placebo (p 5 0.08). The corresponding figures for the patients adhering to the protocol during the first period were 20.017 L versus 20.198 L (p , 0.05, confidence interval of the difference: 20.035 to 10.327 L). For all patients entered, as well as for patients adhering to the trial, there was always a trend in favor of budesonide, as judged by changes in FEV1 and FVC in both 3-mo periods. None of the patients had asthma, but the patients on budesonide had a mean improvement in histamine reactivity of 11.15 dose steps over the entire 6-mo period, as opposed to 10.017 dose steps in patients on placebo (p , 0.05). There was also a significant (p 5 0.01) correlation between pre-trial histamine reactivity and the change in FEV1 in the first period in patients on budesonide. We conclude that inhaled glucocorticosteroids can be of short-term benefit in patients with CF and chronic P.a. infection and that those patients most likely to benefit from this treatment are patients with hyperreactive airways. Prolonged studies in larger number of patients are necessary to determine the long-term efficacy of this treatment. Bisgaard H, Pedersen SS, Nielsen KG, Skov M, Laursen EM, Kronborg G, Reimert CM, Høiby N, Koch C. Controlled trial of inhaled budesonide in patients with cystic fibrosis and chronic bronchopulmonary Pseudomonas aeruginosa infection. AM J RESPIR CRIT CARE MED 1997;156:1190–1196.

Patients with cystic fibrosis (CF) suffer from recurrent and chronic infections which cause inflammatory reactions that have been described as early, severe, and sustained (1). Inflammation is detectable in early infancy as shown by studies of bronchoalveolar lavage (BAL) fluid (2) and high levels of pro-inflammatory cytokines can be found in BAL fluid even in patients that are clinically stable (3). Chronic infection with Pseudomonas aeruginosa (P.a.) is acquired with age by the majority of patients, and the subsequent vigorous host immune response leads to a condition of chronic neutrophil inflammation with progressive decline in lung function (4). The recognition of the damaging role of the host defense mecha-

(Received in original form December 9, 1996 and in revised form June 5, 1997) Supported by Astra Draco. Correspondence and requests for reprints should be addressed to Hans Bisgaard, Department of Pediatrics, National University Hospital, Rigshospitalet, DK-2100 Copenhagen, Denmark. Am J Respir Crit Care Med Vol. 156. pp. 1190–1196, 1997

nism in this situation has led to studies of the effect of antiinflammatory treatment, and the nonsteroidal agent ibuprofen has shown clinical efficacy as judged by short-term changes in lung function (5). Glucocorticosteroids have immunosuppressive as well as anti-inflammatory activity, and alternate-day prednisolone treatment of children with CF has proved beneficial (6), but in a subsequent multi-center study a significantly reduced linear growth, abnormal glucose tolerance, and increased colonization with P.a. was found (7, 8). Inhaled glucocorticosteroids have an improved therapeutic index and might therefore be of value in the treatment of the immune-mediated inflammatory lung disease that characterizes chronic P.a. infection in CF, without having significant systemic side effects. We therefore undertook a controlled single center trial of the short-term effects and side effects of 6 mo of treatment with inhaled glucocorticosteroid in patients with CF and chronic P.a. infection. Outcome variables included changes in lung function, bronchial reactivity, and biochemical inflammatory markers in blood and sputum.

Bisgaard, Pedersen, Nielsen, et al.: Budesonide in CF Patients with P. aeruginosa

METHODS The study was approved by the local ethics committee (L92-238), and all patients gave informed consent. In patients , 18 yr of age, consent was also obtained from their parents.

Patients Patients 8 yr of age and older suffering from CF and chronic P.a. infection were eligible for entry. Chronic P.a. infection was defined as the continuous isolation of P.a. in monthly samples of lower airway secretions for at least 6 mo, or for a shorter period of time but with a significant increase in specific serum anti-P.a. antibodies (9). Patients were excluded if they presented with asthma, allergic bronchopulmonary aspergillosis, colonization with Burkholderia cepacia, or severely impaired cardiac, hepatic, or renal function, if they were pregnant, or if they had been treated with inhaled or systemic glucocorticosteroids within 2 mo prior to entry.

Active Treatment Active treatment consisted of budesonide, 800 mg twice daily, delivered as dry powder inhaled from a Turbuhaler®. Lactose dry powder from a Turbuhaler twice daily was used as placebo.

Study Design The study design was a single-center, double-blind, randomized, controlled, parallel trial. Patients were randomized to treatment groups of equal size in balanced blocks of four patients. Treatment codes were not broken before clean file was declared by the data manager according to GCP recommendations. The study monitor kept track of drug accountability.

Routine Treatment Routine treatment of all patients with chronic P.a. infection in our center includes regular elective courses of 2 wk of intravenously administered antibiotics approximately every third month (9). Between treatment courses, the patients are seen in the out-patient clinic every 4 wk. Trial treatment was started at the end of an intravenous treatment course and was given during two successive intervals, thus covering approximately 6 mo (Figure 1). Each of the two intervals was considered one treatment period. Trial medication was renewed at each clinic visit. Patients were excluded from “per protocol” evaluation of the first treatment period if the interval between the intravenous courses was , 8 wk or . 20 wk. Patients were excluded from “per protocol” evaluation of the whole treatment period if the sum of the two intervals was , 18 wk or . 40 wk.

Other (Routine) Treatment During the trial, all other components of treatment were continued according to routine protocols. These include antibiotics (including

1191

oral and inhaled anti-Pseudomonas antibiotics), PEP-mask physiotherapy, pancreatic enzymes, and inhaled b2-agonists. None of the patients received other topical or systemic glucocorticosteroid treatment or rhDNase inhalations.

Lung Function Lung function was evaluated by spirometry using a Dräger pneumotachograph. b2-agonists were omitted 4 h prior to the test, and the best of three valid attempts was used. Results were compared with control reference standards based on the subject’s height and sex and expressed as percentage of predicted values (10).

Bronchial Reactivity Bronchial reactivity to histamine and to standardized exercise challenge was measured at the end of the intravenous antibiotic courses before and after the trial (Figure 1). Challenge was not performed if the baseline FEV1 was , 1 L.

Histamine Challenge Histamine chloride was nebulized from a Wright nebulizer with a 4-ml fill and a dynamic flow of 8 L/min. Saline was used initially, and the challenge omitted if this caused a change in FEV1 . 10%. The concentration of histamine chloride was doubled until the FEV1 fell by 20% or the maximal concentration of 8 mg/ml was reached. the PC20 was estimated from semi-log interpolation (11).

Exercise Challenge Exercise challenge was performed prior to the histamine challenge, using a treadmill with a 108 inclination, adjusting the speed until a pulse of approximately 180/min was reached, usually within 2 min, after which exercise was continued for a total of 6 min. FEV1 was recorded before and 3, 5, 10, and 15 min after exercise. The response was evaluated from the lowest FEV1 after exercise as compared with the pre-challenge value.

Blood and Sputum Samples Blood and sputum samples were obtained at entry, before each of the two succeeding intravenous courses, and at the end of the trial, as shown in Figure 1. Sputum samples were obtained by expectoration and immediately frozen at 2808 C. The sol phase of thawed samples was obtained by ultracentrifugation at 120,000 3 g for 4 h and subsequently kept frozen at 2808 C until analyzed.

Clinical Chemistry Parameters Clinical chemistry parameters, assessed in blood before and at the end of the trial, included aspartate aminotransferase, alkaline phosphatase, lactate dehydrogenase, bilirubin, total leukocyte and differential count, hemoglobin, thrombocytes, sedimentation rate, C-reactive protein, IgG, IgA, and IgM, albumin, blood urea nitrogen, and creatinine.

Inflammatory Markers

Figure 1. Study scheme.

Sputum sol phase. Tumor necrosis factor (TNF) was measured by ELISA as reported earlier (12). Natural human TNF obtained from lipopolysaccharide-stimulated whole blood from healthy volunteers, as described by Desch and associates (13), was used as standard. The sensitivity of the assay was 30 pg TNF per milliliter. Interleukin-1 receptor antagonist (IL-1ra) was measured by ELISA as described previously (12), using natural human IL-1ra as standard. Polymorphonuclear granulocyte (PMN) elastase was measured by a commercial assay (Merck, Darmstadt, Germany), in which elastase complexed to a 1-proteinase inhibitor is measured by ELISA, using purified human PMN elastase (Sigma Chemical Co., St. Louis, MO) as standard. Free elastolytic activity was measured by a colorimetric assay, using the chromogenic substrate N-methoxysuccinyl-ala-ala-pro-val-pnitroanilide (Sigma). Plasma. Precipitating antibodies against P.a. sonicated whole-cell antigens were measured by crossed immunoelectrophoresis (14), and IgG antibodies by ELISA (15). PMN elastase was measured by the same technique as in sputum sol phase described above.

1192

AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE

VOL. 156

1997

Adrenal Function All patients , 18 yr were subjected to an ACTH stimulation test before and at the end of the trial.

Drug Compliance Drug compliance was estimated by asking the patients to return the Turbuhalers at each clinic visit and counting the remaining number of doses.

Statistical Methods Lung function at the end of the pre-trial intravenous antibiotic course was used as baseline, and treatment effect was evaluated by comparison with lung function at the beginning of the two subsequent intravenous courses (Figure 1). Treatment effects on PC20 (histamine) and on response to exercise were evaluated from the changes from the end of the first to the end of the last intravenous antibiotic course of the trial (Figure 1). The Student’s t test was used for all comparisons. Values of p , 0.05 were considered statistically significant. Mean and SD were used as descriptive statistics.

RESULTS Fifty-five patients were included in the study, and 30 were randomized to active treatment. The patients in the budesonide group were significantly taller (mean height: 170 cm versus 164 cm; p , 0.05), as well as heavier (mean weight: 60 kg versus 51 kg; p , 0.01) than those in the placebo group. However, mean FEV1 (61% versus 59%) and mean FVC (77% versus 75%) showed no significant differences at entry. Two sets of analysis were carried out: analysis of all patients treated (APT) (5 analysis by “intention to treat”) and analysis of “per protocol” treated patients (PPT), i.e., patients who adhered to the protocol, either in the first period only or in both periods. The mean number of days in the study was 96 (range: 62 to 138) for patients completing the first period and 196 (range: 131 to 255) for those completing both periods. Analysis of All Patients Treated (APT; n 5 55)

In the budesonide group, the change in FEV1 from initial value during the first period was 20.032 L versus 20.187 L in the placebo group (p 5 0.08). The change from initial value to end of trial (both periods) was 10.002 L in the budesonide group versus 20.098 L in the placebo group (p . 0.1). The changes in FVC followed the same pattern and the difference between the two groups was always in favor of budesonide for FEV1 as well as for FVC, but between-group comparisons failed to reach statistical significance. Forty-one patients were challenged with histamine and exercise before and after the entire 6-mo treatment. At entry, there were no differences between the two groups in the mean responses to these stimuli. After the 6-mo treatment, the histamine PC20 had increased by 1.15 dose steps in the patients on budesonide (n 5 21) as compared with 0.017 dose steps in the patients on placebo (n 5 19) (Figure 2). This difference (1.13 dose steps; 95% confidence interval: 0.01 to 2.26) between the groups was statistically significant (p , 0.05), showing that budesonide treatment had led to an improvement in bronchial reactivity to histamine. Before treatment, one patient had a decline in FEV1 of 18% in response to exercise, while the rest of the patients had declines of , 12%. However, in contrast to the change in histamine reactivity following budesonide treatment, the response to exercise was unchanged after the 6-mo treatment in both groups.

Figure 2. Bronchial reactivity to histamine before and after the 6-mo trial period. A significantly reduced reactivity was seen in the group of patients treated with budesonide, but there was no change in the group treated with placebo (p , 0.05).

Analysis of “Per Protocol” Treated Patients (PPT; n 5 46 for the First Period Only and n 5 42 for Both Periods)

First period only. Before breaking the study code, nine patients were excluded from evaluation of the first trial period. Six patients, three in each treatment group, required a premature intravenous course, thus not allowing for the 8-wk trial treatment required, and were therefore considered treatment failures. Three patients were withdrawn because of noncompliance, one in the active group and two in the placebo group. The changes in FEV1 in the remaining 46 patients completing the first period were 20.017 L in patients on budesonide versus 20.198 L in those on placebo. The difference between groups was statistically significant (p , 0.05; 95% confidence interval of the difference: 10.035 L to 10.327 L) (Figure 3). The changes in FVC showed the same trend as for FEV1 but the difference between the two groups did not reach statistical significance. The change in FEV1 in patients on budesonide in the first period correlated with the pre-trial responsiveness to histamine (coefficient of correlation 5 20.53; p 5 0.01) (Figure 4). Patients on budesonide with low PC20 (high bronchial reactivity) profited more from budesonide treatment as compared with those with high PC20 (low or normal bronchial reactivity). Both treatment periods. Thirteen patients were excluded from evaluation of the entire 6-mo treatment period; four because of treatment failure (two in each treatment group), and eight because of noncompliance (five in the budesonide group and three in the placebo group). One patient from the placebo group was withdrawn in the second treatment period because of an adverse event, a peri-tonsillar abscess. The reduced number of patients completing both treatment periods (n 5 42) again showed a statistically significant effect of budesonide in the first treatment period, but for the

1193


Figure 3. Change in FEV 1 during the first treatment period. A mean fall of 0.198 L was found in the group treated with placebo compared with a fall of 0.017 L in the group of patients treated with budesonide (p , 0.05), as indicated by the bars. The 95% confidence interval for the advantage of budesonide over placebo was 10.035 L to 10.327 L.

whole study period the difference between the two groups was no longer statistically significant. The changes in lung function parameters consistently showed a trend in favor of budesonide, as was also found in analysis of all patients entered. The inflammatory markers exhibited a marked variability at baseline. The coefficient of variation ranged from 44% to 113%. None of the inflammatory markers changed significantly from baseline within or between the two treatment groups. Drug compliance: Not all patients remembered to return the trial medication, but based upon the remaining doses in the inhalers that were returned, we estimated that the overall percentage of trial medication actually taken by the patients during the entire 6-mo period was z 60%, with no difference between the two groups. Adverse events are shown in Table 1. The overall incidences of reported adverse events were comparable, i.e., no single adverse event was more prevalent in the group of active treatment. Fifteen patients were subjected to an ACTH test. All except two were normal, as reflected by a 30-min plasma cortisol value . 0.5 nmol/L or an increase of 0.19 nmol/L. One patient in each treatment group exhibited an abnormal response, but in both cases this was found before the trial and had normalized at the end of trial. The response to ACTH did not differ between treatment groups before or after treatment. Clinical chemistry parameters exhibited no time- or treatment-dependent changes.

DISCUSSION Chronic bronchopulmonary P.a. infection is at present responsible for the majority of excess morbidity and mortality in patients with CF, and although advances have been made by aggressive antimicrobial treatment (17), many patients still reach end-stage respiratory insufficiency in early adulthood. This is

Figure 4. Change in FEV1 during budesonide treatment versus pretrial bronchial reactivity to histamine (coefficient of correlation 5 20.53; p 5 0.01).

the result of years of inflammatory tissue damage; besides antibiotic treatment, several therapeutic approaches have been taken to reduce the inflammation and thereby prevent tissue damage. In the original placebo-controlled study of Auerbach and colleagues (6), a significant effect on respiratory morbidity and a significant reduction in serum IgG was reported in children with CF who received alternate-day prednisolone for 3.5 yr. Treatment was given to patients irrespective of infection with P.a., and side effects were not reported. These findings prompted a multi-center study of alternate-day oral prednisolone in two doses: 2 and 1 mg/kg bodyweight versus placebo in children with an FEV1 . 60% predicted (8). Treatment with 2 mg/kg on alternate days was stopped prematurely because of side effects (8), and although treatment with prednisolone 1 mg/kg on alternate days resulted in a significantly higher FVC and FEV1 than placebo, it also led to significantly reduced linear growth, abnormal glucose metabolism, and in-

TABLE 1 AU: PLEASE PROVIDE TITLE FOR TABLE

Dysphonia Moniliasis Hemoptysis Cough Pneumothorax Chest pain Abscess Pharyngitis Pneumonia Infection/fever Total Visits, total

Budesonide

Placebo

6 4 2 — — 1 — 3 1 —

3 2 — 2 1 — 1 — 1 3

17

13

174

146

1194


creased rate of colonization with P.a. (7). Greally and coworkers (18), however, found no effect on lung function after 12 wk of treatment with prednisolone, initially 2 mg/kg/d for 2 wk followed by alternate-day prednisolone 1 mg/kg for a further 10 wk. The use of oral glucocorticosteroids is hampered by their unfavorable therapeutic index, but this can be improved by topical application (19). Budesonide is an inhaled glucocorticosteroid of high topical activity and low systemic activity, and when it is delivered as a dry-powder from a Turbuhaler, the therapeutic index is further improved because of higher lung deposition (20). High-dose budesonide Turbuhaler is therefore likely to lead to efficient local effect of glucocorticosteroid with minimal systemic side effects. Because lung function declines with increasing inflammation, we used a design where the drug was given between courses of intravenous antibiotics and where the primary end-point was the change in lung function during two such successive periods of approximately 3 mo each. This is a design that has previously been used to demonstrate the beneficial effect of inhaled colistin (21) and it is readily applied to our patients, who are treated with regular courses of intravenous antibiotics every 3 to 4 mo (17). Treatment is therefore initiated at time of maximal lung function, which then gradually declines between courses, as a consequence of increasing bacterial growth leading to increasing inflammation and airflow obstruction (22). In the reduced number of 46 patients who actually completed the first 3-mo period, the decline in FEV1 was significantly less in the budesonide group than in the placebo group (Figure 3). The relatively limited fall in lung function in the placebo group is probably due to the extensive routine treatment protocol applied in our patients, including regular inhalation of colistin (21). The apparently stable lung function in those on topical steroid indicates that the glucocorticosteroid has impeded the progressive inflammation that is normally seen between courses and that is the usual consequence of gradual proliferation of P.a. in the bronchial tree when intravenous antibiotics are stopped (17, 22). Only 42 patients completed the entire 6-mo trial period and the low number may explain why the effect of budesonide was now no longer significant, since there was a consistent trend that budesonide was superior to placebo and since the same trend was seen when we performed the “intention to treat” analysis on all patients entered into the trial (APT). None of the outcome variables showed a trend in favor of placebo. Theoretically, the high drop-out number may have biased the outcome if the drop-out was differential, but there was an equal drop-out number in the two groups. The less-than-optimal drug compliance may also have had an impact on outcome, but the fact that significant efficacy of budesonide was demonstrable in the first period only, where the number of patients completing the trial was highest, suggests that the decreased number of patients completing both periods has led to a type II error. Adverse events in the budesonide-treated group of patients were mild and not different from those in the placebo group (Table 1). In an early study from our center, we used beclomethasone dipropionate 400 mg/d for 4 mo in 26 patients with chronic P.a. infection (13 on active and 13 on placebo) without observing any significant difference (23). However, the dose may have been too low to be effective or the patient number too small. Nikolaizik and Schöni (24) gave 30 d of inhaled beclomethasone dipropionate, 1,500 mg/d in an open, placebo-controlled study in CF patients with chronic P.a. infection during admission to hospital for antibiotic treatment. No differences were found in FEV1 and FVC changes, possibly because the effect

VOL. 156

1997

of intravenous antibiotic treatment overshadowed any effect from added glucocorticosteroid treatment. During a 2-yr study of inhaled fluticasone propionate in adult patients with CF, Nieman and associates (25) found that active treatment showed some clinical benefit. Their study was also limited by a high drop-out rate, as only 17 of 36 patients enrolled could be analyzed after 2 yr. Atopy appears to be more common among patients with CF (26, 27), and the presence of asthma might confound any apparent treatment effect from glucocorticosteroids. Standardized exercise challenge was therefore performed to disclose concurrent asthma, since exercise rather than methacholine provocation has previously been recommended to distinguish asthma from other chronic lung diseases, including CF (28). Only one patient had a decrease in FEV1 of 18% before the trial, while no other patients showed a decline . 12% either before or after the trial. Also, there were no significant changes in response to exercise in the budesonide-treated group, making it highly unlikely that the treatment effect could be ascribed to a treatment of concurrent bronchial asthma. Many patients with CF have bronchial hyperreactivity (29– 31), and it is not known whether this is a reflection of inflammatory changes exposing irritant receptors, as is the prevailing concept of the pathophysiology in bronchial asthma, or if the bronchi in patients with CF have an abnormal autonomic response (32–36). Patients with CF could also be hyperreactive simply by virtue of the reduced airway caliber secondary to inflammatory changes, which predisposes to an exaggerated increase in airway resistance due to the inverse fourth-power ratio between change in airway diameter and resistance. In any case, bronchial reactivity may be considered a marker of airway pathophysiology in CF and appears to be an unfavorable prognostic finding, as demonstrated in a prospective study in which the natural history of lung diseases was more severe in hyperresponsive patients with CF (30). The present study showed a decrease in bronchial hyperreactivity in response to glucocorticosteroid treatment as judged by the response to histamine challenge. Though the magnitude of improvement is hardly of clinical significance in itself, decreased bronchial reactivity may reflect an ameliorating effect of glucocorticosteroids on airway pathophysiology. As a surrogate marker, it may serve as an indicator of an improvement in the long-term course of the lung disease. Within the active treatment group, the change in FEV1 during the first 3 mo of treatment correlated with the baseline PC20 (histamine), indicating that the more hyperreactive patients derived more benefit from the treatment. This correlation would seem to further substantiate an anti-inflammatory effect of the budesonide treatment. Local and systemic levels of a number of cytokines such as TNF, IL-1, IL-2, and IL-8 are frequently increased in patients with CF (12, 37–41). We did not find any effect of budesonide treatment on sputum levels of TNF and the IL-1 receptor antagonist, even in the first treatment period, where the fall in FEV1 was significantly less in the budesonide group, but it is likely that sputum samples are much less representative of local concentrations in the bronchial tree than BAL fluid. The dominating effector cells in the bronchial lumen in infected patients with CF are neutrophils, and increased local and circulating levels of PMN elastase are found (37) and are thought to be a marker of neutrophil activation and thus disease severity (42). However, in the patients treated as per protocol (PPT), we could not detect any difference in elastase between the budesonide and placebo group, either in sputum or in plasma. The highly fluctuating pattern of these markers, combined with our observation that the baseline coefficient of


variation of TNF, IL-1ra, and PMN elastase ranged from 44% to 113% in this study, would increase the risk of a type II error. Eosinophilic cationic protein is released from activated eosinophils and increased concentrations can be found in serum and sputum from infected patients with CF, implying that eosinophils are actively involved in the CF lungs (43). We found no changes in eosinophilic cationic protein in the present study, implying that an effect of budesonide upon activation of eosinophils was unlikely in the present study. There were no detectable changes in levels of P.a. precipitins or IgG antibodies as measured by ELISA, but levels of these antibodies are relatively stable in individual patients over shorter periods of time (9). Reviewing the various trials of systemic and topical glucocorticosteroid treatment of patients with CF, it seems that treatment does have an effect on CF lung disease. The present study seems to add “proof of concept”; however, the magnitude of efficacy is rather limited and it is doubtful whether the present data support the widespread use of inhaled glucocorticosteroids in CF. Those most likely to benefit from it seem to be patients with hyperreactive airways. The principle of anti-inflammatory treatment or immunomodulation by inhibiting pro-inflammatory mediators needs to be explored even further. The encouraging results of the study of treatment with the nonsteroidal anti-inflammatory drug ibuprofen for 4 yr (5) supports the rationale of this principle. The results of the present single-center, short-term, pilot study may serve as the impetus for larger, multi-center, longterm studies of topical glucocorticosteroid treatment of CF patients with chronic P.a. infection. Since clinical and laboratory evidence indicate that inflammatory reactions in response to viral and/or bacterial infections are present already in the first year of life of most patients with CF, and since inflammatory tissue damage might further secondary colonization with pathogens, clinical trials of inhaled glucocorticosteroids in infants and young children with CF should be considered. This will be possible when techniques for noninvasive measurement of lung function indices in awake children become available. References 1. Cantin, A. 1995. Cystic fibrosis lung inflammation: early, sustained and severe. Am. J. Respir. Crit. Care Med. 151:939–940. 2. Khan, T. Z., J. S. Wagener, T. Boat, J. Martinez, F. J. Accurso, and D. W. H. Riches. 1995. Early pulmonary inflammation in infants with cystic fibrosis. Am. J. Respir. Crit. Care Med. 151:1075–1082. 3. Konstan, M. W., K. A. Hilliard, T. M. Norvell, and M. Berger. 1994. Bronchoalveolar lavage findings in cystic fibrosis patients with stable, clinically mild lung disease suggest ongoing infection and inflammation. Am. J. Respir. Crit. Care Med. 150:448–454. 4. Koch, C., and N. Høiby. 1993. Pathogenesis of cystic fibrosis. Lancet 341: 1065–1069. 5. Konstan, M. W., P. J. Byard, C. L. Hoppel, and P. B. Davies. 1995. Effect of high-dose ibuprofen in patients with cystic fibrosis. N. Engl. J. Med. 332:848–854. 6. Auerbach, H. S., M. Williams, J. A. Kirkpatrick, and H. R. Colten. 1985. Alternate-day prednisone reduces morbidity and improves pulmonary function in cystic fibrosis. Lancet 2:686–688. 7. Eigen, H., B. J. Rosenstein, S. FitzSimmons, and D. V. Schidlow. 1995. A multicenter study of alternate-day prednisone therapy in patients with cystic fibrosis. J. Pediatr. 126:515–523. 8. Rosenstein, B. J., and H. Eigen. 1991. Risks of alternate-day prednisone in patients with cystic fibrosis. Pediatrics 87:245–246. 9. Høiby, N., E. W. Flensborg, B. Beck, B. Friis, S. V. Jacobsen, and L. Jacobsen. 1977. Pseudomonas aeruginosa infection in cystic fibrosis: diagnostic and prognostic significance of Pseudomonas aeruginosa precipitins determined by means of crossed immunoelectrophoresis. Scand. J. Respir. Dis. 58:65–79. 10. Sherrill, D. L., M. D. Lebowitz, R. J. Knudson, and B. Burrows. 1992.

11.

12.

13.

14.

15.

16.

17.

18.

19. 20. 21.

22.

23.

24.

25.

26.

27. 28.

29.

30.

31. 32.

33. 34.

1195

Continuous longitudinal regression equations for pulmonary function measures. Eur. Respir. J. 5:452–462. Cockcroft, D. W., K. Y. Murdock, and J. T. Mink. 1983. Determination of histamine PC20: comparison of linear and logarithmic interpolation. Chest 84:505–506. Kronborg, G., M. B. Hansen, M. Svenson, A. Fomsgaard, N. Høiby, and K. Bendtzen. 1993. Cytokines in sputum and serum from patients with cystic fibrosis and chronic Pseudomonas aeruginosa infection as markers of destructive inflammation in the lungs. Pediatr. Pulmonol. 15: 292–297. Desch, C. E., N. L. Kovach, W. Present, C. Bryles, and J. M. Harlan. 1989. Production of human tumour necrosis factor from whole blood ex vivo. Lymphokine Res. 8:141–146. Høiby, N. 1977. Pseudomonas aeruginosa infection in cystic fibrosis: diagnostic and prognostic significance of Pseudomonas aeruginosa precipitins determined by crossed immunoelectrophoresis: a survey. Acta Pathol. Microbiol. Scand. 262(Suppl.):1–96. Pedersen, S. S., F. Espersen, and N. Høiby. 1987. Diagnosis of chronic Pseudomonas aeruginosa infection in cystic fibrosis by enzyme-linked immunosorbent assay. J. Clin. Microbiol. 25:1830–1836. Reimert, C. M., P. Venge, A. Kharazmi, and K. Bendtzen. 1991. Detection of eosinophilic cationic protein (ECP) by an enzyme-linked immunosorbent assay. J. Immunol. Methods 138:285–290. Frederiksen, B., S. Lanng, C. Koch, and N. Høiby. 1996. Improved survival in the Danish center-treated cystic fibrosis patients: results of aggressive treatment. Pediatr. Pulmonol. 21:153–158. Greally, P., M. J. Hussain, D. Vergani, and J. F. Price. 1994. Interleukin1a, soluble interleukin-2 receptor, and IgG concentrations in cystic fibrosis treated with prednisolone. Arch. Dis. Child. 71:35–39. Toogood, J. H. 1989. High-dose inhaled steroid therapy for asthma. J. Allergy Clin. Immunol. 83:528–536. Lipworth, B. J. 1995. New perspectives on inhaled drug delivery and systemic bioactivity. Thorax 50:105–110. Jensen, T., S. S. Pedersen, S. Garne, C. Heilmann, N. Høiby, and C. Koch. 1987. Colistin inhalation therapy in cystic fibrosis with chronic Pseudomonas aeruginosa lung infection. J. Antimicrob. Chemother. 19: 831–838. Regelmann, W. E., G. R. Elliott, J. Warwick, and C. C. Clawson. 1990. Reduction of sputum Pseudomonas aeruginosa density by antibiotics improves lung function in cystic fibrosis more than do bronchodilators and chest physiotherapy alone. Am. Rev. Respir. Dis. 141:914–921. Schiotz, P. O., M. Jørgensen, E. W. Flensborg, O. Faerø, S. Husby, N. Høiby, S. V. Jacobsen, H. Nielsen, and S. E. Svehag. 1983. Chronic Peudomonas aeruginosa lung infection in cystic fibrosis. Acta Paediatr. Scand. 72:283–287. Nikolaizik, W. H., and M. H. Schoni. 1996. Pilot study to assess the effect of inhaled corticosteroids on lung function in patients with cystic fibrosis. J. Pediatr. 128:271–274. Nieman, R., S. Williams, C. Maden, R. Knight, and M. Hodson. 1996. A double blind, placebo-controlled study comparing the effects of the inhaled corticosteroid fluticasone dipropionate 500 mg bid with placebo in adults with cystic fibrosis (abstract). Am. J. Respir. Crit. Care Med. 153:A72. Tobin, M. J., O. Maguire, D. Reen, E. Tempany, and M. X. Fitzgerald. 1980. Atopy and bronchial reactivity in older patients with cystic fibrosis. Thorax 35:807–813. 1979. Allergy and bronchial hyperreactivity in cystic fibrosis (editorial). Lancet i:708. Godfrey, S., C. Springer, N. Noviski, C. H. Maayan, and A. Avital. 1991. Exercise but not methacholine differentiates asthma from chronic lung disease in children. Thorax 46:488–492. Mitchell, I., M. Corey, R. Woenne, I. R. Krastins, and H. Levison. 1978. Bronchial hyperreactivity in cystic fibrosis and asthma. J. Pediatr. 93: 744–748. Eggleston, P. A., B. J. Rosenstein, C. M. Stackhouse, and M. F. Alexader. 1988. Airway hyperreactivity in cystic fibrosis: clinical correlates and possible effects on the course of the disease. Chest 94:360–365. Cropp, G. J. 1996. Effectiveness of bronchodilators in cystic fibrosis. Am. J. Med. 100(Suppl. 1A):19s–29s. Davis, P. B., J. R. Shelhamer, and M. Kaliner. 1980. Abnormal adrenergic and cholinergic sensitivity in cystic fibrosis. N. Engl. J. Med. 302: 1453–1456. Davis, P. B. 1984. Autonomic and airway reactivity in obligate heterozygotes for cystic fibrosis. Am. Rev. Respir. Dis. 129:911–914. van Asperen, P., C. M. Mellis, and R. T. South, and S. J. Simpson. 1981. Bronchial reactivity in cystic fibrosis with normal pulmonary function. Am. J. Dis. Child. 135:815–819.

1196


35. van Asperen, P. P., P. Manglick, and H. Allen. 1988. Mechanisms of bronchial hyperreactivity in cystic fibrosis. Pediatr. Pulmonol. 5:139– 144. 36. van Haren, E. H. J., J. W. Lammers, J. Festen, and C. L. van Herwaarden. 1992. Bronchial vagal tone and responsiveness to histamine, exercise and bronchodilators in adult patients with cystic fibrosis. Eur. Respir. J. 5:1083–1088. 37. Suter, S., U. B. Schaad, P. Roux-Lombard, E. Girardin, G. Grau, and J.-M. Dayer. 1989. Relation between tumor necrosis factor-alfa and granulocyte elastase-alfa-1 proteinase inhibitor complexes in the plasma of patients with cystic fibrosis. Am. Rev. Respir. Dis. 140:1640–1644. 38. Wilmott, R. W., J. T. Kassab, P. L. Kilian, W. R. Benjamin, S. D. Douglas, and R. E. Wood. 1990. Increased levels of interleukin-1 in broncho-alveolar washings from children with bacterial pulmonary infections. Am. Rev. Respir. Dis. 142:365–368. 39. Kronborg, G., A. Fomsgaard, G. H. Shand, S. Lanng, M. B. Hansen, T.

40.

41.

42.

43.

VOL. 156

1997

Pressler, and N. Høiby. 1992. TNF-1a and immune complexes in sputum and serum from patients with cystic fibrosis and chronic Pseudomonas aeruginosa lung infection. Immunol. Infect. Dis. 2:171–177. Greally, P., M. J. Hussein, A. J. Cook, A. P. Sampson, P. J. Piper, and J. F. Price. 1993. Sputum tumour necrosis factor-alfa and leukotriene concentrations in cystic fibrosis. Arch. Dis. Child. 68:389–392. Dean, T. P., Y. Dai, J. K. Shute, M. K. Church, and J. O. Warner. 1993. Interleukin-8 concentrations are elevated in bronchoalveolar lavage, sputum, and sera of children with cystic fibrosis. Pediatr. Res. 34:159– 161. Fick, R. B. 1993. Pathogenetic mechanisms for cystic fibrosis lung disease: a paradigm for inflammatory airways disease. J. Lab. Clin. Med. 121:632–634. Koller, D. Y., M. Götz, I. Eichler, and R. Urbanek. 1994. Eosinophilic activation in cystic fibrosis. Thorax 49:496–499.