|
|
| |||||||||||||||||||||||||||||||||||||||
Background Direct aerosol delivery of aminoglycosides such as tobramycin to the lower airways of patients with cystic fibrosis may control infection with Pseudomonas aeruginosa and improve pulmonary function, with low systemic toxicity. We conducted a randomized crossover study to evaluate the safety and efficacy of aerosolized tobramycin in patients with cystic fibrosis and P. aeruginosa infections.
Methods Seventy-one patients with stable pulmonary status were recruited from seven U.S. centers for the treatment of cystic fibrosis and randomly assigned to one of two crossover regimens. Group 1 received 600 mg of aerosolized tobramycin for 28 days, followed by half-strength physiologic saline (placebo) for two 28-day periods. Group 2 received placebo for 28 days, followed by tobramycin for two 28-day periods. Pulmonary function, the density of P. aeruginosa in sputum, ototoxicity, nephrotoxicity, and the emergence of tobramycin-resistant P. aeruginosa were monitored.
Results In the first 28-day period, treatment with tobramycin was associated with an increase in the percentage of the value predicted for forced expiratory volume in one second (9.7 percentage points higher than the value for placebo; P<0.001), forced vital capacity (6.2 percentage points higher than the value for placebo; P = 0.014), and forced expiratory flow at the midportion of the vital capacity (13.0 percentage points higher than the value for placebo; P<0.001). A decrease in the density of P. aeruginosa in sputum by a factor of 100 (P<0.001) was found during all periods of tobramycin administration. Neither ototoxicity nor nephrotoxicity was detected. The frequency of the emergence of tobramycin-resistant bacteria was similar during both tobramycin and placebo administration.
Conclusions The short-term aerosol administration of a high dose of tobramycin in patients with clinically stable cystic fibrosis is an efficacious and safe treatment for endobronchial infection with P. aeruginosa.
Previous clinical trials evaluating the efficacy of aerosolized aminoglycosides in patients with cystic fibrosis have yielded conflicting results4. Four studies noted a decrease in the rates of hospitalization5,6,7,8. Several studies demonstrated that the use of the drugs led to either a significant improvement5,6,7,9 or a slower decline10,11 in pulmonary function, as compared with placebo. Other studies noted no change in pulmonary function8,12. The interpretation of these results was limited by the smallness of the samples,6,7,8,9,12 a lack of placebo-control groups,7,8,10,13,14 inadequate masking of the taste of the antibiotic, and a failure to consider potential carryover effects in crossover designs5,6,9,11,12,15. In addition, there were marked differences in the amount of aminoglycoside placed in the nebulizer (ranging from 20 mg11 to 600 mg15,16) and the type of nebulizer used to deliver aerosolized antibiotics to the airways5,6,7,8,9,10,11,12,13,14,15,16,17.
To overcome some of the limitations of previous studies, we conducted a multicenter, double-blind, placebo-controlled, three-period crossover trial to determine the safety and efficacy of aerosolized tobramycin for the treatment of endobronchial infection due to P. aeruginosa in patients with cystic fibrosis.
Methods
Patients
Between March 1989 and June 1991, 71 patients were enrolled at seven U.S. centers treating cystic fibrosis. All patients had a diagnosis of cystic fibrosis confirmed by a sweat test,18 sputum containing P. aeruginosa susceptible to tobramycin (zone diameter, 
16 mm after testing with a 10-µg tobramycin disk), and a forced vital capacity (FVC) of 40 percent or more of the value predicted for height. Patients were excluded if their sputum yielded pseudomonas species other than P. aeruginosa, their auditory threshold in either ear was above 20 dB at frequencies between 500 and 8000 Hz, their serum creatinine concentration was 1.5 mg per deciliter or higher ( 133 µmol per liter), or their blood urea nitrogen concentration was above 25 mg per deciliter (>9 mmol per liter). Patients were enrolled when their FVC was within 10 percent of the best value obtained within the previous six months, or two weeks after they completed antibiotic therapy for a pulmonary exacerbation (i.e., while pulmonary status was considered to be stable).
Informed consent, as approved by the human-subjects review board at each center, was obtained from all patients at enrollment.
Regimens
All patients received two inhalations of metaproterenol (0.65 mg per inhalation) by metered-dose inhaler 15 minutes before each administration of drug or placebo. With a nose clip in place, they then inhaled aerosol containing either placebo (30 ml of half-strength physiologic saline) or tobramycin three times a day by ultrasonic nebulizer (Ultraneb 100/99, DeVilbiss, Somerset, Pa.)16,19. All patients completed 200 tidal inhalations. Previous in vitro data20 had indicated that tobramycin concentrations in sputum had to be more than 10 times the minimal inhibitory concentration of tobramycin-susceptible P. aeruginosa (i.e., 400 µg per gram of sputum) to ensure that the bacteria in the sputum were killed. Reaching this target concentration of tobramycin required the administration of 600 mg of preservative-free tobramycin sulfate dissolved in 30 ml of half-strength physiologic saline (adjusted to a pH of 6.85 to 7.05), a dose greater than those used in previous regimens5,7,10,11,12. Quinine hydrochloride (1 mg per milliliter) was added to the saline excipient for both drug and placebo administration, to mask the taste of the tobramycin and to serve as a measure of compliance with the protocol.
Each patient received both placebo and tobramycin in a crossover fashion, after being randomly assigned to one of two groups treated in the following sequences: group 1 received tobramycin for 28 days, followed by placebo for two 28-day periods; group 2 received placebo for 28 days, followed by tobramycin for two 28-day periods. Randomization was stratified within each center according to the severity of illness at enrollment (FVC <70 percent of the predicted value vs. FVC 70 percent), by means of a stratified block design21.
Monitoring
The patients were evaluated at enrollment (visit 1), during weeks 2, 4, 8, and 12 of drug or placebo administration (at visits 2, 3, 4, and 5, respectively), and four weeks after administration (at visit 6). The level of pulmonary function, the density of bacteria in sputum, and the complete blood count were determined at each visit.
At visits 1 and 5, each patient was examined by a physician and a disease-status score was recorded22. This score reflects the severity of illness and uses a 100-point scale to assess 12 categories of pulmonary, nutritional, and general health status.
To monitor nephrotoxicity and ototoxicity, serum creatinine measurements, urinalysis, and testing of auditory acuity and vestibular function were performed during visits 1, 3, 4, and 5; auditory acuity was also tested 28 weeks after the administration of the study agents. Acuity was measured in both ears at frequencies between 500 and 8000 Hz,23 and vestibular function was assessed with the dynamic E test24.
All sputum samples were shipped on ice to the core laboratory (Children's Hospital and Medical Center, Seattle) for quantitative culture (described below) and the determination of tobramycin susceptibility at each visit. To expedite enrollment, one half of the initial sample was cultured at the participating study center for the presence of P. aeruginosa and susceptibility to tobramycin, according to standard clinical methods25. Only microbiologic findings in samples cultured at the core laboratory were used in the data analysis.
The FVC, the forced expiratory volume in one second (FEV1), the forced expiratory flow at the midportion of vital capacity (FEF25-75%), and the total lung capacity (TLC) were recorded and expressed as percentages of the predicted values (the standards were adapted from those of Knudsen et al26). Residual volume (RV) and the ratio of RV to TLC were calculated. Pulmonary-function testing followed published guidelines for assessment in children27 and adults28. The results were expressed as percentages predicted from standards specific for age, sex, and height (adapted from Knudsen et al26). Absolute lung volumes were measured by volume-displacement plethysmography27.
During each 28-day period of drug or placebo administration, the patients were contacted on random dates and asked to mail a urine sample to the core laboratory. The urine was assayed for its quinine concentration, as a measure of compliance (see below).
A pulmonary exacerbation was indicated by at least two of the following seven symptoms during the study: fever (oral temperature, >38 °C), more frequent coughing (increase of 50 percent), increased sputum volume (increase of 50 percent), loss of appetite, weight loss of at least 1 kg, absence from school or work (at least three of the preceding seven days) due to illness, and symptoms of an upper respiratory tract infection. These symptoms had to have been associated with at least one of three additional criteria: a decrease in the FVC of at least 10 percent, an increase in the respiratory rate of at least 10 breaths per minute, and a peripheral-blood neutrophil count of 15,000 per cubic millimeter or more.
Bacteriologic Procedures
Sputum specimens were cultured with a quantitative technique29 modified to increase the probability of isolating P. cepacia30. All P. aeruginosa morphotypes were tested for their susceptibility to tobramycin by a semiautomated broth-dilution method31,32. Tobramycin resistance was indicated by a minimal inhibitory concentration of 8 µg per milliliter or more32.
Quinine Assay
The urinary quinine concentration was measured by high-pressure liquid chromatography33. Curves for urinary quinine washout in 16 patients with cystic fibrosis who were not enrolled in the current study showed that the mean (±SE) time required for the excretion of 90 percent of the inhaled quinine (formulated as in this study) was 40.2 ±12.3 hours. When 90 percent had been excreted, the urinary quinine concentration had decreased to the limit of detection by the chromatographic method, 100 ng per milliliter. According to these data, noncompliance was indicated by a random urine sample with a quinine concentration below 100 ng per milliliter.
Statistical Analysis
The study had a three-period crossover design. The use of a third period provided greater power for detecting carryover effects34,35.
We performed two evaluations of the efficacy of 28-day treatment: one evaluation was based on all three periods and the crossover design, and the other was based only on the first period and a parallel design. The latter evaluation was performed because of concern about the increased use of antibiotics (other than tobramycin) among patients who received placebo during the second and third study periods. Two-sample t-tests were used to compare mean differences in outcome between the two study groups in the first-period parallel analysis. For the three-period crossover analysis, F tests were calculated according to a mixed-effects model in which overall mean, period, treatment, and carryover terms were considered fixed effects and subject and error terms were considered random effects. Potential confounders and covariates such as compliance, antibiotic use, base-line pulmonary status, and center effects were investigated with multiple regression methods for the parallel analysis and with generalized estimating equations36 for the dependent variables in the crossover analysis. Comparisons of proportions were tested with the chi-square test or Fisher's exact test when appropriate. When necessary, patients were randomly excluded to balance the sizes of the groups as required for the crossover analysis. Patients were included in analyses on the basis of their group assigned at enrollment (intention-to-treat analysis).
For the primary outcome measures -- FEV1, FVC, and FEF25-75% -- treatment-effect estimates were considered to be significant if P was 0.016 or less (with Bonferroni's adjustment for multiple comparisons). P values for secondary outcomes are reported, but because of multiple comparisons they should be interpreted only as relative measures of an association between outcome and treatment.
Results
Thirty-six patients were assigned to group 1, and 35 to group 2. There were no significant differences between the two study groups in age, sex, base-line FVC and FEV1 (percent of the predicted values), or density of P. aeruginosa at base line (P>0.25 for all comparisons) (Table 1). The mean disease-status score22 at enrollment was higher in group 1 than in group 2 (P = 0.02).
|
The frequency of noncompliance with medication (estimated from the urinary quinine level) ranged from 7 to 20 percent across the three periods. Compliance was comparable during the periods of tobramycin and placebo administration (P>0.35 for all three periods).
Outcome
Pulmonary-Function Tests
The mean improvement in the FVC, FEV1, and FEF25-75% from enrollment to the end of the first study period (day 28) was greater during tobramycin administration than during placebo administration, according to parallel analysis (P<0.015 for all three measures) (Figure 1 and Table 2). Pulmonary function, expressed as a percentage of the predicted values, improved by 6 to 13 percentage points across the three measures. The mean change in the RV/TLC ratio from base line was 5.5 percentage points less (i.e., indicating improvement) during tobramycin administration than during placebo administration (P = 0.013).
|
|
|
A decrease in the density of P. aeruginosa in sputum by a factor of 100 was associated with tobramycin administration (Figure 2). This change was observed in both the first-period parallel analysis (P<0.001) (Table 2) and the three-period crossover analysis (P<0.001) (Table 3) without detectable carryover or period effects. The total peripheral-blood white-cell count was lower during tobramycin administration than during placebo administration (P = 0.029 for the first-period analysis; P = 0.089 for the three-period analysis). The mean peripheral-blood polymorphonuclear neutrophil count was also lower during tobramycin administration according to both the parallel and the crossover analyses (P<0.02).
|
During the first and second study periods there were no differences between the two study groups in the frequency of pulmonary exacerbations (P>0.4) (Table 4). By the third period, fewer patients had pulmonary exacerbations during tobramycin administration (3 percent) than during placebo administration (20 percent) (P = 0.06). A more pronounced trend was observed in the use of antibiotics (Table 4); by the third period, more patients (49 percent) received antibiotic treatment (oral and parenteral) initiated during placebo administration than during tobramycin administration (15 percent) (P = 0.006).
|
The serum creatinine concentrations of all patients remained in the normal range throughout the observation period. They rose transiently by 50 percent in six patients during tobramycin administration and in five patients during placebo administration. Cellular casts were not detected in the urine of any patient during the observation period. Proteinuria ( 2+) was detected in two patients (one receiving tobramycin and one receiving placebo) on one visit, but not at the subsequent visit.
No clinically important or statistically significant change occurred in auditory acuity in either study group at any frequency between 500 and 8000 Hz tested during the 42-week observation period. Ototoxicity (a decrease of 20 dB in auditory acuity at any frequency in either ear after enrollment) and symptoms of vestibular dysfunction did not occur in any patient, and tests of vestibular function were negative in all patients.
Emergence of Tobramycin-Resistant Pseudomonas Species
The clinical laboratories of the study centers did not isolate tobramycin-resistant P. aeruginosa from any sputum samples obtained at enrollment; the core laboratory, however, cultured these organisms from the samples of two patients. Resistant P. aeruginosa were found in sputum from 10 of 71 patients (14 percent) during the monitoring period (i.e., by visit 6). The frequency of the emergence of resistant strains was similar during tobramycin administration and placebo administration (P>0.5).
The emergence of P. cepacia and P. maltophilia (Xanthomonas maltophilia) was also monitored. At enrollment, P. cepacia was found in sputum from two patients and P. maltophilia in one patient according to the core laboratory, although these pathogens had not been identified by the laboratories of the study centers. In addition, 3 patients became infected with P. cepacia during the study period, and 10 with P. maltophilia. There was no significant difference between the frequency of the emergence of either species during tobramycin administration and the frequency during placebo administration (P>0.7).
Discussion
The aerosol administration of 600 mg of tobramycin in the first 28-day treatment period was associated with an improvement in the FVC, FEV1, and FEF25-75%, as well as a decrease in air trapping (reflected by the RV/TLC ratio), as compared with placebo administration. These findings are consistent with several previous studies5,6,7,9. Although these changes are statistically significant, they are small. Accompanying these changes in pulmonary function was a substantial decrease in the density of P. aeruginosa in sputum and a smaller decrease in the peripheral-blood neutrophil count. Contrary to earlier findings,7,16 the treatment effect of tobramycin was not influenced by the severity of illness.
The most pronounced improvement in pulmonary function occurred in the first 28 days of tobramycin administration. In a previous study16 we observed a decline in the efficacy of tobramycin when given longer than 28 days. In an attempt to determine the possible cause of this apparent decline in treatment effect, we undertook an extensive analysis of potential confounding factors, including the patients' adherence to the treatment protocol, their pulmonary function at enrollment, and the emergence of P. aeruginosa strains resistant to tobramycin. The only trend we observed was an increase in antibiotic use among patients who received placebo in the third period, as compared with the patients who received tobramycin. This increased use of antibiotics during placebo administration could have reduced the relative difference in treatment effect between the two study groups. It is also possible that continued administration of tobramycin after 28 days would maintain but not increase the treatment effect.
The short-term administration of high doses of tobramycin by ultrasonic nebulizer appears to be safe. Measurements of auditory, vestibular, and renal function remained within the normal range in both study groups, a finding consistent with previous results16. Concern has been expressed about the emergence of tobramycin-resistant strains of P. aeruginosa10,14,16. The frequency of the emergence of these organisms, as well as the emergence of P. cepacia and P. maltophilia intrinsically resistant to tobramycin, was not increased during tobramycin administration relative to placebo administration. Whether longer-term administration would increase the frequency of colonization by tobramycin-resistant bacteria is not known.
The intravenous administration of antibiotics is the currently accepted treatment for pulmonary exacerbations of cystic fibrosis in patients with endobronchial infection due to P. aeruginosa1. After 14 days of administration of intravenous antibiotics, the FEV1 increases 20 percent37 and the density of P. aeruginosa in sputum decreases substantially38,39. It is difficult to compare the results of the present study with those of studies of intravenous antibiotic administration, since the study agents were administered to the patients in this study while their pulmonary status was stable, rather than during a pulmonary exacerbation. This difference in respiratory status may explain the smaller increase in the FEV1 in our study (9 percentage points) as compared with that in the study of intravenous antibiotic administration (20 percentage points).
In the patients with stable respiratory status, tobramycin was more effective than placebo in improving pulmonary function and decreasing the density of P. aeruginosa for four weeks. By prolonging optimal pulmonary status in patients with cystic fibrosis, this drug may decrease the frequency of courses of intravenous antibiotic treatment. The cost of administering aerosolized tobramycin three times a day is one-half the daily cost of administering two intravenous antibiotics at home and one-fifth the daily cost of administering two intravenous antibiotics in hospitals in the Seattle area. The ongoing development of more efficient, inspiration-activated (rather than continuous-flow) nebulizers will greatly decrease the loss of tobramycin during aerosol administration, thus decreasing the dose (the amount placed in the nebulizer) and the cost of the antibiotic.
Supported by a grant (A021 9-1) from the Cystic Fibrosis Foundation.
We are indebted to Dr. Richard Lemen (Department of Pediatrics, University of Arizona, Tucson) for his contributions to the study design and implementation, as well as his review of the manuscript; to Judy Williams-Warren and Sharon McNamara for their supervision of the study implementation, as well as their review of the manuscript; to Donna Hedges, the data manager; to Marty Cohen, the research pharmacist; to Allan Weber, the laboratory supervisor; to research coordinators Karin Coyne (San Diego, Calif.), Kathy Hyman (Philadelphia), Vikki Kociela (Cincinnati), Jane McNamara (Boston), Sara Rae (Portland, Oreg.), and Michelle Todd (Tucson), for their contributions to the implementation of the study protocol; to the patients and their families for their participation in the study; and to Tanya Desloover for assistance in the preparation of the manuscript.
Source Information
From the Departments of Pediatrics (B.W.R., R.L.G., S.J.A., K.W., A.L.S.), Biostatistics (M.A.M.), and Epidemiology (S.J.A.), Schools of Medicine and Public Health and Community Medicine, University of Washington, and the Children's Hospital and Medical Center, both in Seattle; the Department of Pediatrics, Tufts University School of Medicine, Boston (H.L.D.); the Department of Pediatrics, Oregon Health Sciences University, Portland (J.D.E.); the Department of Pediatrics, University of California School of Medicine, San Diego (I.R.H.); the Department of Pediatrics, Temple University School of Medicine, Philadelphia (D.V.S.); and the Department of Pediatrics, University of Cincinnati, Cincinnati (R.M.K., R.W.W.).
Address reprint requests to Dr. Ramsey at the Cystic Fibrosis Program, Children's Hospital and Medical Center, 4800 Sand Point Way N.E., P.O. Box C-5371, Seattle, WA 98105.
References
| |||||||||||||||||||||||||||||||||||||||
Related Letters:
Aerosolized Tobramycin in Patients with Cystic Fibrosis
Ashby B. L., Ramsey B. W., Smith A. L.
Extract |
Full Text
N Engl J Med 1993;
329:1659-1660, Nov 25, 1993.
Correspondence
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2008 Massachusetts Medical Society. All rights reserved. |
Previous
