- Email: [email protected]
Administration of Aerosolized Antibiotics in Cystic Fibrosis Patients
Administration of Aerosolized Antibiotics in Cystic Fibrosis Patients* Richard B. Moss, MD, FCCP
High rates of colonization and the challenge of managing Pseudomonas aeruginosa infections in patients with cystic fibrosis (CF) have necessitated a search for safe and effective antibiotics. Currently, therapy with an aminoglycoside in combination with a -lactam or a quinolone antibiotic is the standard. Unfortunately, it is difficult to deliver high doses of these antibiotics via the IV route without significant systemic adverse events (AEs) (eg, ototoxicity and nephrotoxicity). Recently, a reformulation of the aminoglycoside antibiotic tobramycin has become available in a preservative-free, pH-adjusted solution for inhalation by jet nebulizer. A 96-week series of clinical studies including 520 patients, aged > 6 years, with moderate-to-severe CF has evaluated the long-term safety and effectiveness of this formulation. Patients received tobramycin solution for inhalation (TSI) or placebo, which was administered in alternating cycles of 28-days-on and 28-days-off therapy, plus their usual CF care for 6 months with open-label follow-up extended to 2 years. Most AEs declined in frequency with increasing TSI exposure. Patients receiving TSI spent 25 to 33% fewer days in the hospital. Following the initiation of TSI treatment, patients experienced significant increases in FEV1. FEV1 values were maintained above baseline for the duration of the study series. Antibiotic susceptibility of the bacterial isolates did not predict clinical response. TSI was safe, well-tolerated, and effective for long-term treatment (96 weeks) of P aeruginosa colonization and infection in CF patients. (CHEST 2001; 120:107S–113S) Key words: aerosolized antibiotics; cystic fibrosis; Pseudomonas aeruginosa; tobramycin solution for inhalation Abbreviations: AE ⫽ adverse event; CF ⫽ cystic fibrosis; MIC ⫽ minimal inhibitory concentration; TSI ⫽ tobramycin solution for inhalation
feature of cystic fibrosis (CF) lung disease is A major endobronchial infection with the bacterium Pseudo-
monas aeruginosa. Early in their lives, CF patients may be infected with Staphylococcus aureus or Haemophilus influenzae, but by the age of 17 years, nearly 70% of CF patients have P aeruginosa present in their sputum cultures.1 P aeruginosa, which can be present at densities of *From the Department of Pediatric Pulmonary Medicine, Stanford University Medical Center, Palo Alto, CA. The author is a paid consultant for PathoGenesis Corporation (Seattle, WA), Genentech, Inc (South San Francisco, CA), and AeroGen, Inc (Sunnyvale, CA). Correspondence to: Richard B. Moss, MD, FCCP, Pediatric Pulmonary Medicine, Stanford University Medical Center, 701 Welch Rd, Room 3328, Palo Alto, CA, CA 94304-5786; e-mail: [email protected]
106 to 108 colony-forming units per gram of sputum, is the major infectious burden in the airway of CF patients.2 Once established in CF patients, respiratory tract infections are not eradicated by antibiotic therapy. Respiratory disease in CF patients is characterized by the progressive obstruction of the airways and loss of lung function due in large part to inflammatory response to chronic bacterial infection. The loss of pulmonary function resulting in respiratory failure is a primary cause of death in CF patients. In the 1998 Cystic Fibrosis Patient Registry, 87.4% of registry deaths could be attributed to the loss of pulmonary function.3 The presence of P aeruginosa is associated with increased rates of lung function decline4 and is a significant independent predictor of mortality.5 The role of chronic inflammation in the pulmonary function decline of CF patients has been confirmed by separate reports on the efficacy of long-term anti-inflammatory agent therapy (ie, ibuprofen6 and prednisone7) in slowing lung function decline. CF patients are prone to episodes of acute pulmonary exacerbation, characterized by worsening symptoms of respiratory tract infection accompanied by acute declines in lung function. The primary cause of CF patient hospitalization is for the treatment of exacerbations, with ⬎ 35% of patients hospitalized at least once annually.1 The central role of P aeruginosa in CF lung disease has led to the testing of intensive therapy with antipseudomonal antibiotics to suppress infection,8 even in the absence of pulmonary exacerbations. Among patients with well-established infections, the suppression of P aeruginosa has been shown to lead to decreases in sputum P aeruginosa density. Although these decreases can be short-lived beyond the cessation of therapy,9 lung function benefits from antibiotic therapy can be maintained over extended periods.
Use of Tobramycin Tobramycin is the most frequently prescribed aminoglycoside for the treatment of pulmonary infections in CF patients. Tobramycin is frequently administered IV during periods of acute exacerbations. Because the penetration of tobramycin into sputum is low following IV administration, high doses are required to achieve concentrations inhibitory to P aeruginosa. Moreover, the inactivation of tobramycin in purulent sputum mandates the delivery of 10 to 25 times the minimal inhibitory concentration (MIC) to achieve bacterial killing.10 However, high doses of IV tobramycin increase the risk of systemic adverse events (AEs) such as ototoxicity and nephrotoxicity.
Aerosolized Tobramycin Recently, a nonpyrogenic, preservative-free, pH-adjusted formulation of tobramycin solution for inhalation (TSI) (TOBI; Chiron Corporation; Emeryville, CA), administered via a jet nebulizer, has become commercially available for use in CF patients. This product allows the delivery of the antibiotic directly to the endobronchial space in the lungs, while minimizing systemic exposure and the associated risk of ototoxicity and nephrotoxicity. CHEST / 120 / 3 / SEPTEMBER, 2001 SUPPLEMENT
107S
TSI Study Series The long-term safety and efficacy of TSI has been evaluated in a 96-week series involving 520 CF patients. The results of two double-blind, placebo-controlled trials representing the first 24 weeks of this series have been published previously.11
Patient Criteria The selection criteria for the pivotal studies were chosen in order to enroll CF patients who have P aeruginosa and moderate-to-severe obstructive airway disease, but who were otherwise in relatively stable condition at enrollment. Only CF patients ⬎ 6 years of age who had an FEV1 between 25% and 75% of predicted values and a sputum or throat culture yielding P aeruginosa were eligible for enrollment. Patients were excluded from the study if they had used any antipseudomonal antibiotic within 14 days of receiving their initial dose of the study drug. Patients also were excluded from the study if they had undergone a respiratory culture positive for Burkholderia cepacia in the previous 2 years, had compromised renal function, had hypersensitivity to aminoglycosides, had a recent episode of hemoptysis, or were pregnant at screening.
CF (ie, usual care), except for any inhaled antibiotics other than TSI. Figure 1 illustrates the two phases. In the randomized phase, patients were treated with either TSI, 300 mg bid, or taste-masked placebo bid. In the open-label phase, all patients received TSI, 300 mg bid. The study drug was administered twice daily using a jet nebulizer (PARI LC PLUS; PARI Respiratory Equipment, Inc; Richmond, VA) and a compressor (PulmoAid; DeVilbiss; Somerset, PA). The overall design of each study is illustrated in Figure 2. Each study consisted of three drug administration cycles. A cycle included a 4-week on-drug period followed by a 4-week off-drug period. Thus, each study included three 4-week on-drug periods and three 4-week off-drug periods. The total length of this study series was 96 weeks. Five hundred twenty patients were enrolled in the randomized phase of the study series.11 Four hundred sixty-four patients completed the randomized phase. Of the 464 patients who completed the randomized phase, 396 entered the open-label phase. A total of 242 patients completed all three of the studies in the open-label phase for a total of up to 96 weeks of intermittent TSI therapy.
Patient Demographics Table 1 summarizes the patient demographics at the start of the randomized and open-label phases.
Study Design This study series consisted of a randomized pivotal phase and an open-label phase. The pivotal phase included two identical, double-blind, placebo-controlled studies, each of 24 weeks’ duration. The open-label phase included three sequential 24-week studies. Throughout the series, patients could receive any and all standard therapies for
Analysis The safety and efficacy data from this study series have been examined in two ways. First, ordinal data, such as those from pulmonary function tests, can be analyzed by original treatment group using the start of the randomized
Figure 1. Study series phases. TOBI ⫽ TSI. 108S
History and Evolution of Aerosolized Therapeutics
Figure 2. Study design.
phase as a reference (eg, the percentage change in FEV1 percent predicted from baseline). This approach allows statistical analysis of the change from baseline. Second, categoric data such as AEs may be evaluated in an intent-to-treat analysis in which patients who received placebo in the pivotal studies and subsequently switched to TSI are grouped with those patients who received TSI from the start of the series. This approach allows evaluation of the occurrence of events as a function of time receiving TSI. In such analyses, time may be represented as discrete 3-cycle (12-week) blocks of exposure or by the number of weeks of exposure, with the first dose of TSI as the baseline.
AEs The incidence of most AEs remained constant or decreased with increasing TSI exposure. Four adverse experiences (ie, vomiting, fever, abdominal pain, and anorexia) were reported by significantly more patients receiving placebo than those receiving TSI (p ⱕ 0.05)
Table 1—Patient Demographics at the Start of the Randomized and Pivotal Phases Start of Randomized Phase (n ⫽ 520) Demographics Male gender* Age, yr† FEV1, % predicted†
TSI (n ⫽ 258) 149 (58) 20.8 (9.5) 50 (15.5)
*Values given as No. (%). †Values given as mean (SD).
Start of OpenLabel Phase (n ⫽ 396)
Placebo (n ⫽ 262) 132 (50) 20.6 (10.0) 51 (16.8)
211 (53) 20 (9.7) 52 (15.7)
during the pivotal studies (Fig 3).11 As the length of TSI exposure increased, the rates of these events declined further. The decrease in incidence of these adverse experiences may reflect an improvement in some of the chronic symptoms of CF with ongoing TSI therapy. As reported by Ramsey et al,11 two AEs that were reported by significantly (p ⱕ 0.05) more TSI patients than placebo patients during the pivotal studies were voice alteration and tinnitus (Fig 3). In the open-label phase, the incidence of voice alteration decreased steadily with increasing TSI exposure (from 9.1 to 3.9%). Voice alteration was usually mild to moderate in severity and tended to improve during the off-drug part of each cycle. Although there was a significant difference in the incidence of tinnitus between treatment groups in the randomized part of the series, tinnitus was rare among TSI-treated patients. The frequency of tinnitus never exceeded 3.5% during the study series (Fig 3). During the study series, a total of 14 patients (12 adults and 2 adolescents) reported tinnitus. Most episodes of tinnitus were transient and mild or moderate in severity. Audiology testing showed that none of the 14 patients reporting tinnitus had objective evidence of hearing loss, which was defined as a bilateral decrease from baseline in hearing thresholds of ⱖ 15 decibels at two consecutive frequencies.
Effect on Renal Function Mean serum creatinine levels were comparable between the TSI and placebo groups at the start of the randomized phase. No clinically significant changes in creatinine levels occurred over the course of the study series. The mean values were within normal limits at all observations.
Colonization by Other Organisms Long-term antibiotic therapy may alter the composition of the lung flora and lead to changes in the rates at which CHEST / 120 / 3 / SEPTEMBER, 2001 SUPPLEMENT
109S
Figure 3. Adverse experiences with significant differences between treatment groups in the pivotal trials.
other potential pathogens are isolated.12 Furthermore, the odds of isolating rarely occurring organisms have been increased by improvements in microbiological techniques that allow for the more accurate identification of organisms. Thus, an accurate assessment of the effect of TSI on the lung flora in CF patients is critical. Microbiological data from this study series were analyzed to evaluate whether long-term TSI treatment was associated with increased isolation of Gram-negative organisms that are intrinsically resistant to tobramycin (ie, Burkholderia cepacia, Stenotrophomonas maltophilia, or Alcaligenes xylosoxidans) or the fungus Aspergillus. The incidence of isolation of B cepacia did not increase during up to 12 cycles (96 weeks) of TSI treatment. The incidence of isolation of both S maltophilia and A xylosoxidans was shown to increase over time, but the rate of increase in isolation frequency was the same or greater
during 24 weeks of placebo exposure, and, thus, these increases may be independent of TSI exposure. The incidence of Aspergillus isolation increased with increasing TSI exposure. No increase in the rate of Aspergillus isolation was observed during the 24-week placebo exposure. The clinical significance of the increase in Aspergillus isolation is not known, but no increase in investigatordefined allergic bronchopulmonary aspergillosis was found in 24 months of observation.
Other Outcomes Measures Hospitalization Hospitalization is a highly costly and disruptive aspect of the traditional management of CF. During the initial randomized, controlled trial, patients receiving placebo
Figure 4. Percentage of days hospitalized per season. * ⫽ all hospitalizations; LRTD ⫽ lower respiratory disease. 110S
History and Evolution of Aerosolized Therapeutics
Figure 5. Courses of IV antipseudomonal therapy per patient per year.
spent an average of 8 days in the hospital. Following the initiation of TSI therapy, patients spent 25 to 33% fewer days in the hospital, suggesting a beneficial effect on morbidity. Hospitalization rates for CF patients are seasonal, with significantly increased rates of hospital admission in autumn and winter when viral superinfections are most common. Rates tend to decrease in spring and summer. An analysis of the entire 96-week study series demonstrated that in each season TSI-treated patients were hospitalized for a smaller percentage of days than that observed in patients during 24 weeks of placebo exposure. The decreases in the percentage of days hospitalized in each season of the first and second years of the study series were comparable (Fig 4).
IV Antipseudomonal Antibiotic Use An alternative means of assessing patient outcome is by evaluation of the use of additional forms of antipseudomonal therapy during treatment. Following the initiation of TSI therapy, the number of patients requiring IV antibiotic therapy was reduced by 20 to 25% relative to that observed during 24 weeks of placebo exposure in the pivotal trials (Fig 5). Furthermore, the use of oral quinolone therapy was reduced approximately 33% over the entire study series.
Change in Pulmonary Function Change in pulmonary function is usually considered to be a primary and clinically important outcome measure in
Figure 6. Mean relative change in FEV1 percent predicted by week.
CHEST / 120 / 3 / SEPTEMBER, 2001 SUPPLEMENT
111S
Figure 7. Change in FEV1 percent predicted by P aeruginosa isolate tobramycin susceptibility.
CF patients. In the randomized phase, TSI-treated patients had a sharp increase in FEV1 values following the start of TSI therapy (Fig 6). By the end of the last on-drug period of the randomized phase (ie, after 20 weeks of therapy), a treatment effect of 11.9% in FEV1 values was observed between TSI-treated patients and placebotreated patients.11 Following the initiation of TSI treatment, patients originally treated with placebo also exhibited an improvement in FEV1 values. In both original treatment groups, FEV1 was maintained above baseline from the initiation of TSI treatment to the end of the study series. Improvements in FEV1 following the initiation of TSI treatment and long-term maintenance of this improvement clearly demonstrate the efficacy of TSI therapy.
Microbiological Response The clinical response to TSI therapy cannot be predicted by P aeruginosa isolate susceptibility. At the end of the randomized phase, patients with isolates in the higher MIC categories (ie, ⬎ 8 g/mL) had smaller mean relative changes in FEV1 percent predicted than those patients whose highest MIC isolates were in the ⱕ 8-g/mL category (Fig 7). However, despite having isolates with tobramycin MIC values between 16 and 64 g/mL, a similar percentage of patients had improved FEV1 values when compared with those patients whose isolates had MIC values of ⱕ 8 g/mL. The mean relative change in FEV1 percent predicted appeared to be independent of the tobramycin MIC at all other points in the series. The mean improvement in FEV1 percent predicted and the proportion of patients with positive clinical responses were similar in all three MIC categories. These findings suggest that the conventional definition of drug resistance and sensitivity (ie, MIC breakpoint of 8 to 16 g/mL) as determined by parenteral tobramycin 112S
therapy may not be relevant for TSI. Rather, due to the high sputum levels achieved with TSI, a new definition of susceptibility may need to be developed. Apparently, modest decreases in P aeruginosa susceptibility are not predictive of clinical efficacy for TSI for a 2-year period.
Subgroup Analysis Treatment randomization was prospectively stratified using a number of factors that permitted subgroup analysis. These factors included the use of dornase alfa (Pulmozyme; Genentech; South San Francisco, CA), disease severity, gender, and age. At the end of the randomized phase (week 24), TSI patients in the age groups of patients 13 to 17 years and ⱖ 18 years had significantly (pⱕ 0.05) greater relative changes in FEV1 than did placebo patients.13 Although FEV1 percent predicted for TSI patients in the age group of patients 6 to 12 years was improved substantially, the p value was 0.076 due to improvements in the placebo-treated patients in this age group. Furthermore, TSI patients in all subgroups defined by gender, disease severity, and rhDNAse use had significant (pⱕ 0.05) improvements in FEV1 percent predicted relative to placebo-treated patients in these subgroups.
Summary Similar to the results of the 6-month randomized, placebo-controlled trial with TSI therapy,11 the long-term (24-month) treatment data from the open-label extension study demonstrated the following conclusions about TSI therapy: 1. it provided sustained improvements in pulmonary function; 2. it reduced the duration and frequency of hospitalization; History and Evolution of Aerosolized Therapeutics
3. it reduced the need for concomitant antipseudomonal therapy; and 4. it is safe, effective, and well-tolerated in patients with CF.
7
In addition, the 24-month study also showed that pathogen susceptibility was not a predictor of clinical effect.
8
References
9
1 Cystic Fibrosis Foundation. Patient registry: 1997 annual data report. Bethesda, MD: Cystic Fibrosis Foundation, 1998 2 Davis PB, Drumm M, Konstan MW. Cystic fibrosis. Am J Respir Crit Care Med 1996; 154:1229 –1256 3 Cystic Fibrosis Foundation. Patient registry: 1998 annual data report. Bethesda, MD: Cystic Fibrosis Foundation, 1999 4 Pamukcu A, Bush A, Buchdahl R. Effects of P aeruginosa colonization on lung function and anthropomorphic variables in children with cystic fibrosis. Pediatr Pulmonol 1995; 19:10 –15 5 Henry RL, Mellis CM, Petrovic L. Mucoid Pseudomonas aeruginosa is a marker of poor survival in cystic fibrosis. Pediatr Pulmonol 1992; 12:158 –161 6 Konstan MW, Byard PJ, Hoppel CL, et al. Effect of high-dose
10
11
12
13
ibuprofen in patients with cystic fibrosis. N Engl J Med 1995; 332:848 – 854 Eigen H, Rosenstein BJ, FitzSimmons S, et al. A multicenter study of alternate-day prednisone therapy in patients with cystic fibrosis. J Pediatr 1995; 126:515–523 Szaff M, Hoiby N, Flensberg EW. Frequent antibiotic therapy improves survival of cystic fibrosis patients with chronic Pseudomonas aeruginosa infection. Acta Pediatr Scand 1983; 72:651– 657 Cullen RT, McCrae WM, Govan J, et al. Ceftazidime in cystic fibrosis: clinical, microbiological and immunological studies. J Antimicrob Chem 1983; 12:369 –375 Mendelman PM, Smith AL, Levy J, et al. Aminoglycoside penetration, inactivation, and efficacy in cystic fibrosis sputum. Am Rev Respir Dis 1985; 132:761–765 Ramsey BW, Pepe MS, Quan JM, et al. Intermittent administration of inhaled tobramycin in patients with cystic fibrosis. N Engl J Med 1999; 340:23–30 Denton M, Todd NJ, Littlewood JM. Role of anti-pseudomonal antibiotics in the emergence of Stenotrophomonas maltophilia in cystic fibrosis patients. Eur J Clin Microbiol Infect Dis 1996; 15:402– 404 Moss RB. Long-term benefits of inhaled tobramycin in adolescent patients with cystic fibrosis. Chest (in press)
CHEST / 120 / 3 / SEPTEMBER, 2001 SUPPLEMENT
113S
High rates of colonization and the challenge of managing Pseudomonas aeruginosa infections in patients with cystic fibrosis (CF) have necessitated a search for safe and effective antibiotics. Currently, therapy with an aminoglycoside in combination with a -lactam or a quinolone antibiotic is the standard. Unfortunately, it is difficult to deliver high doses of these antibiotics via the IV route without significant systemic adverse events (AEs) (eg, ototoxicity and nephrotoxicity). Recently, a reformulation of the aminoglycoside antibiotic tobramycin has become available in a preservative-free, pH-adjusted solution for inhalation by jet nebulizer. A 96-week series of clinical studies including 520 patients, aged > 6 years, with moderate-to-severe CF has evaluated the long-term safety and effectiveness of this formulation. Patients received tobramycin solution for inhalation (TSI) or placebo, which was administered in alternating cycles of 28-days-on and 28-days-off therapy, plus their usual CF care for 6 months with open-label follow-up extended to 2 years. Most AEs declined in frequency with increasing TSI exposure. Patients receiving TSI spent 25 to 33% fewer days in the hospital. Following the initiation of TSI treatment, patients experienced significant increases in FEV1. FEV1 values were maintained above baseline for the duration of the study series. Antibiotic susceptibility of the bacterial isolates did not predict clinical response. TSI was safe, well-tolerated, and effective for long-term treatment (96 weeks) of P aeruginosa colonization and infection in CF patients. (CHEST 2001; 120:107S–113S) Key words: aerosolized antibiotics; cystic fibrosis; Pseudomonas aeruginosa; tobramycin solution for inhalation Abbreviations: AE ⫽ adverse event; CF ⫽ cystic fibrosis; MIC ⫽ minimal inhibitory concentration; TSI ⫽ tobramycin solution for inhalation
feature of cystic fibrosis (CF) lung disease is A major endobronchial infection with the bacterium Pseudo-
monas aeruginosa. Early in their lives, CF patients may be infected with Staphylococcus aureus or Haemophilus influenzae, but by the age of 17 years, nearly 70% of CF patients have P aeruginosa present in their sputum cultures.1 P aeruginosa, which can be present at densities of *From the Department of Pediatric Pulmonary Medicine, Stanford University Medical Center, Palo Alto, CA. The author is a paid consultant for PathoGenesis Corporation (Seattle, WA), Genentech, Inc (South San Francisco, CA), and AeroGen, Inc (Sunnyvale, CA). Correspondence to: Richard B. Moss, MD, FCCP, Pediatric Pulmonary Medicine, Stanford University Medical Center, 701 Welch Rd, Room 3328, Palo Alto, CA, CA 94304-5786; e-mail: [email protected]
106 to 108 colony-forming units per gram of sputum, is the major infectious burden in the airway of CF patients.2 Once established in CF patients, respiratory tract infections are not eradicated by antibiotic therapy. Respiratory disease in CF patients is characterized by the progressive obstruction of the airways and loss of lung function due in large part to inflammatory response to chronic bacterial infection. The loss of pulmonary function resulting in respiratory failure is a primary cause of death in CF patients. In the 1998 Cystic Fibrosis Patient Registry, 87.4% of registry deaths could be attributed to the loss of pulmonary function.3 The presence of P aeruginosa is associated with increased rates of lung function decline4 and is a significant independent predictor of mortality.5 The role of chronic inflammation in the pulmonary function decline of CF patients has been confirmed by separate reports on the efficacy of long-term anti-inflammatory agent therapy (ie, ibuprofen6 and prednisone7) in slowing lung function decline. CF patients are prone to episodes of acute pulmonary exacerbation, characterized by worsening symptoms of respiratory tract infection accompanied by acute declines in lung function. The primary cause of CF patient hospitalization is for the treatment of exacerbations, with ⬎ 35% of patients hospitalized at least once annually.1 The central role of P aeruginosa in CF lung disease has led to the testing of intensive therapy with antipseudomonal antibiotics to suppress infection,8 even in the absence of pulmonary exacerbations. Among patients with well-established infections, the suppression of P aeruginosa has been shown to lead to decreases in sputum P aeruginosa density. Although these decreases can be short-lived beyond the cessation of therapy,9 lung function benefits from antibiotic therapy can be maintained over extended periods.
Use of Tobramycin Tobramycin is the most frequently prescribed aminoglycoside for the treatment of pulmonary infections in CF patients. Tobramycin is frequently administered IV during periods of acute exacerbations. Because the penetration of tobramycin into sputum is low following IV administration, high doses are required to achieve concentrations inhibitory to P aeruginosa. Moreover, the inactivation of tobramycin in purulent sputum mandates the delivery of 10 to 25 times the minimal inhibitory concentration (MIC) to achieve bacterial killing.10 However, high doses of IV tobramycin increase the risk of systemic adverse events (AEs) such as ototoxicity and nephrotoxicity.
Aerosolized Tobramycin Recently, a nonpyrogenic, preservative-free, pH-adjusted formulation of tobramycin solution for inhalation (TSI) (TOBI; Chiron Corporation; Emeryville, CA), administered via a jet nebulizer, has become commercially available for use in CF patients. This product allows the delivery of the antibiotic directly to the endobronchial space in the lungs, while minimizing systemic exposure and the associated risk of ototoxicity and nephrotoxicity. CHEST / 120 / 3 / SEPTEMBER, 2001 SUPPLEMENT
107S
TSI Study Series The long-term safety and efficacy of TSI has been evaluated in a 96-week series involving 520 CF patients. The results of two double-blind, placebo-controlled trials representing the first 24 weeks of this series have been published previously.11
Patient Criteria The selection criteria for the pivotal studies were chosen in order to enroll CF patients who have P aeruginosa and moderate-to-severe obstructive airway disease, but who were otherwise in relatively stable condition at enrollment. Only CF patients ⬎ 6 years of age who had an FEV1 between 25% and 75% of predicted values and a sputum or throat culture yielding P aeruginosa were eligible for enrollment. Patients were excluded from the study if they had used any antipseudomonal antibiotic within 14 days of receiving their initial dose of the study drug. Patients also were excluded from the study if they had undergone a respiratory culture positive for Burkholderia cepacia in the previous 2 years, had compromised renal function, had hypersensitivity to aminoglycosides, had a recent episode of hemoptysis, or were pregnant at screening.
CF (ie, usual care), except for any inhaled antibiotics other than TSI. Figure 1 illustrates the two phases. In the randomized phase, patients were treated with either TSI, 300 mg bid, or taste-masked placebo bid. In the open-label phase, all patients received TSI, 300 mg bid. The study drug was administered twice daily using a jet nebulizer (PARI LC PLUS; PARI Respiratory Equipment, Inc; Richmond, VA) and a compressor (PulmoAid; DeVilbiss; Somerset, PA). The overall design of each study is illustrated in Figure 2. Each study consisted of three drug administration cycles. A cycle included a 4-week on-drug period followed by a 4-week off-drug period. Thus, each study included three 4-week on-drug periods and three 4-week off-drug periods. The total length of this study series was 96 weeks. Five hundred twenty patients were enrolled in the randomized phase of the study series.11 Four hundred sixty-four patients completed the randomized phase. Of the 464 patients who completed the randomized phase, 396 entered the open-label phase. A total of 242 patients completed all three of the studies in the open-label phase for a total of up to 96 weeks of intermittent TSI therapy.
Patient Demographics Table 1 summarizes the patient demographics at the start of the randomized and open-label phases.
Study Design This study series consisted of a randomized pivotal phase and an open-label phase. The pivotal phase included two identical, double-blind, placebo-controlled studies, each of 24 weeks’ duration. The open-label phase included three sequential 24-week studies. Throughout the series, patients could receive any and all standard therapies for
Analysis The safety and efficacy data from this study series have been examined in two ways. First, ordinal data, such as those from pulmonary function tests, can be analyzed by original treatment group using the start of the randomized
Figure 1. Study series phases. TOBI ⫽ TSI. 108S
History and Evolution of Aerosolized Therapeutics
Figure 2. Study design.
phase as a reference (eg, the percentage change in FEV1 percent predicted from baseline). This approach allows statistical analysis of the change from baseline. Second, categoric data such as AEs may be evaluated in an intent-to-treat analysis in which patients who received placebo in the pivotal studies and subsequently switched to TSI are grouped with those patients who received TSI from the start of the series. This approach allows evaluation of the occurrence of events as a function of time receiving TSI. In such analyses, time may be represented as discrete 3-cycle (12-week) blocks of exposure or by the number of weeks of exposure, with the first dose of TSI as the baseline.
AEs The incidence of most AEs remained constant or decreased with increasing TSI exposure. Four adverse experiences (ie, vomiting, fever, abdominal pain, and anorexia) were reported by significantly more patients receiving placebo than those receiving TSI (p ⱕ 0.05)
Table 1—Patient Demographics at the Start of the Randomized and Pivotal Phases Start of Randomized Phase (n ⫽ 520) Demographics Male gender* Age, yr† FEV1, % predicted†
TSI (n ⫽ 258) 149 (58) 20.8 (9.5) 50 (15.5)
*Values given as No. (%). †Values given as mean (SD).
Start of OpenLabel Phase (n ⫽ 396)
Placebo (n ⫽ 262) 132 (50) 20.6 (10.0) 51 (16.8)
211 (53) 20 (9.7) 52 (15.7)
during the pivotal studies (Fig 3).11 As the length of TSI exposure increased, the rates of these events declined further. The decrease in incidence of these adverse experiences may reflect an improvement in some of the chronic symptoms of CF with ongoing TSI therapy. As reported by Ramsey et al,11 two AEs that were reported by significantly (p ⱕ 0.05) more TSI patients than placebo patients during the pivotal studies were voice alteration and tinnitus (Fig 3). In the open-label phase, the incidence of voice alteration decreased steadily with increasing TSI exposure (from 9.1 to 3.9%). Voice alteration was usually mild to moderate in severity and tended to improve during the off-drug part of each cycle. Although there was a significant difference in the incidence of tinnitus between treatment groups in the randomized part of the series, tinnitus was rare among TSI-treated patients. The frequency of tinnitus never exceeded 3.5% during the study series (Fig 3). During the study series, a total of 14 patients (12 adults and 2 adolescents) reported tinnitus. Most episodes of tinnitus were transient and mild or moderate in severity. Audiology testing showed that none of the 14 patients reporting tinnitus had objective evidence of hearing loss, which was defined as a bilateral decrease from baseline in hearing thresholds of ⱖ 15 decibels at two consecutive frequencies.
Effect on Renal Function Mean serum creatinine levels were comparable between the TSI and placebo groups at the start of the randomized phase. No clinically significant changes in creatinine levels occurred over the course of the study series. The mean values were within normal limits at all observations.
Colonization by Other Organisms Long-term antibiotic therapy may alter the composition of the lung flora and lead to changes in the rates at which CHEST / 120 / 3 / SEPTEMBER, 2001 SUPPLEMENT
109S
Figure 3. Adverse experiences with significant differences between treatment groups in the pivotal trials.
other potential pathogens are isolated.12 Furthermore, the odds of isolating rarely occurring organisms have been increased by improvements in microbiological techniques that allow for the more accurate identification of organisms. Thus, an accurate assessment of the effect of TSI on the lung flora in CF patients is critical. Microbiological data from this study series were analyzed to evaluate whether long-term TSI treatment was associated with increased isolation of Gram-negative organisms that are intrinsically resistant to tobramycin (ie, Burkholderia cepacia, Stenotrophomonas maltophilia, or Alcaligenes xylosoxidans) or the fungus Aspergillus. The incidence of isolation of B cepacia did not increase during up to 12 cycles (96 weeks) of TSI treatment. The incidence of isolation of both S maltophilia and A xylosoxidans was shown to increase over time, but the rate of increase in isolation frequency was the same or greater
during 24 weeks of placebo exposure, and, thus, these increases may be independent of TSI exposure. The incidence of Aspergillus isolation increased with increasing TSI exposure. No increase in the rate of Aspergillus isolation was observed during the 24-week placebo exposure. The clinical significance of the increase in Aspergillus isolation is not known, but no increase in investigatordefined allergic bronchopulmonary aspergillosis was found in 24 months of observation.
Other Outcomes Measures Hospitalization Hospitalization is a highly costly and disruptive aspect of the traditional management of CF. During the initial randomized, controlled trial, patients receiving placebo
Figure 4. Percentage of days hospitalized per season. * ⫽ all hospitalizations; LRTD ⫽ lower respiratory disease. 110S
History and Evolution of Aerosolized Therapeutics
Figure 5. Courses of IV antipseudomonal therapy per patient per year.
spent an average of 8 days in the hospital. Following the initiation of TSI therapy, patients spent 25 to 33% fewer days in the hospital, suggesting a beneficial effect on morbidity. Hospitalization rates for CF patients are seasonal, with significantly increased rates of hospital admission in autumn and winter when viral superinfections are most common. Rates tend to decrease in spring and summer. An analysis of the entire 96-week study series demonstrated that in each season TSI-treated patients were hospitalized for a smaller percentage of days than that observed in patients during 24 weeks of placebo exposure. The decreases in the percentage of days hospitalized in each season of the first and second years of the study series were comparable (Fig 4).
IV Antipseudomonal Antibiotic Use An alternative means of assessing patient outcome is by evaluation of the use of additional forms of antipseudomonal therapy during treatment. Following the initiation of TSI therapy, the number of patients requiring IV antibiotic therapy was reduced by 20 to 25% relative to that observed during 24 weeks of placebo exposure in the pivotal trials (Fig 5). Furthermore, the use of oral quinolone therapy was reduced approximately 33% over the entire study series.
Change in Pulmonary Function Change in pulmonary function is usually considered to be a primary and clinically important outcome measure in
Figure 6. Mean relative change in FEV1 percent predicted by week.
CHEST / 120 / 3 / SEPTEMBER, 2001 SUPPLEMENT
111S
Figure 7. Change in FEV1 percent predicted by P aeruginosa isolate tobramycin susceptibility.
CF patients. In the randomized phase, TSI-treated patients had a sharp increase in FEV1 values following the start of TSI therapy (Fig 6). By the end of the last on-drug period of the randomized phase (ie, after 20 weeks of therapy), a treatment effect of 11.9% in FEV1 values was observed between TSI-treated patients and placebotreated patients.11 Following the initiation of TSI treatment, patients originally treated with placebo also exhibited an improvement in FEV1 values. In both original treatment groups, FEV1 was maintained above baseline from the initiation of TSI treatment to the end of the study series. Improvements in FEV1 following the initiation of TSI treatment and long-term maintenance of this improvement clearly demonstrate the efficacy of TSI therapy.
Microbiological Response The clinical response to TSI therapy cannot be predicted by P aeruginosa isolate susceptibility. At the end of the randomized phase, patients with isolates in the higher MIC categories (ie, ⬎ 8 g/mL) had smaller mean relative changes in FEV1 percent predicted than those patients whose highest MIC isolates were in the ⱕ 8-g/mL category (Fig 7). However, despite having isolates with tobramycin MIC values between 16 and 64 g/mL, a similar percentage of patients had improved FEV1 values when compared with those patients whose isolates had MIC values of ⱕ 8 g/mL. The mean relative change in FEV1 percent predicted appeared to be independent of the tobramycin MIC at all other points in the series. The mean improvement in FEV1 percent predicted and the proportion of patients with positive clinical responses were similar in all three MIC categories. These findings suggest that the conventional definition of drug resistance and sensitivity (ie, MIC breakpoint of 8 to 16 g/mL) as determined by parenteral tobramycin 112S
therapy may not be relevant for TSI. Rather, due to the high sputum levels achieved with TSI, a new definition of susceptibility may need to be developed. Apparently, modest decreases in P aeruginosa susceptibility are not predictive of clinical efficacy for TSI for a 2-year period.
Subgroup Analysis Treatment randomization was prospectively stratified using a number of factors that permitted subgroup analysis. These factors included the use of dornase alfa (Pulmozyme; Genentech; South San Francisco, CA), disease severity, gender, and age. At the end of the randomized phase (week 24), TSI patients in the age groups of patients 13 to 17 years and ⱖ 18 years had significantly (pⱕ 0.05) greater relative changes in FEV1 than did placebo patients.13 Although FEV1 percent predicted for TSI patients in the age group of patients 6 to 12 years was improved substantially, the p value was 0.076 due to improvements in the placebo-treated patients in this age group. Furthermore, TSI patients in all subgroups defined by gender, disease severity, and rhDNAse use had significant (pⱕ 0.05) improvements in FEV1 percent predicted relative to placebo-treated patients in these subgroups.
Summary Similar to the results of the 6-month randomized, placebo-controlled trial with TSI therapy,11 the long-term (24-month) treatment data from the open-label extension study demonstrated the following conclusions about TSI therapy: 1. it provided sustained improvements in pulmonary function; 2. it reduced the duration and frequency of hospitalization; History and Evolution of Aerosolized Therapeutics
3. it reduced the need for concomitant antipseudomonal therapy; and 4. it is safe, effective, and well-tolerated in patients with CF.
7
In addition, the 24-month study also showed that pathogen susceptibility was not a predictor of clinical effect.
8
References
9
1 Cystic Fibrosis Foundation. Patient registry: 1997 annual data report. Bethesda, MD: Cystic Fibrosis Foundation, 1998 2 Davis PB, Drumm M, Konstan MW. Cystic fibrosis. Am J Respir Crit Care Med 1996; 154:1229 –1256 3 Cystic Fibrosis Foundation. Patient registry: 1998 annual data report. Bethesda, MD: Cystic Fibrosis Foundation, 1999 4 Pamukcu A, Bush A, Buchdahl R. Effects of P aeruginosa colonization on lung function and anthropomorphic variables in children with cystic fibrosis. Pediatr Pulmonol 1995; 19:10 –15 5 Henry RL, Mellis CM, Petrovic L. Mucoid Pseudomonas aeruginosa is a marker of poor survival in cystic fibrosis. Pediatr Pulmonol 1992; 12:158 –161 6 Konstan MW, Byard PJ, Hoppel CL, et al. Effect of high-dose
10
11
12
13
ibuprofen in patients with cystic fibrosis. N Engl J Med 1995; 332:848 – 854 Eigen H, Rosenstein BJ, FitzSimmons S, et al. A multicenter study of alternate-day prednisone therapy in patients with cystic fibrosis. J Pediatr 1995; 126:515–523 Szaff M, Hoiby N, Flensberg EW. Frequent antibiotic therapy improves survival of cystic fibrosis patients with chronic Pseudomonas aeruginosa infection. Acta Pediatr Scand 1983; 72:651– 657 Cullen RT, McCrae WM, Govan J, et al. Ceftazidime in cystic fibrosis: clinical, microbiological and immunological studies. J Antimicrob Chem 1983; 12:369 –375 Mendelman PM, Smith AL, Levy J, et al. Aminoglycoside penetration, inactivation, and efficacy in cystic fibrosis sputum. Am Rev Respir Dis 1985; 132:761–765 Ramsey BW, Pepe MS, Quan JM, et al. Intermittent administration of inhaled tobramycin in patients with cystic fibrosis. N Engl J Med 1999; 340:23–30 Denton M, Todd NJ, Littlewood JM. Role of anti-pseudomonal antibiotics in the emergence of Stenotrophomonas maltophilia in cystic fibrosis patients. Eur J Clin Microbiol Infect Dis 1996; 15:402– 404 Moss RB. Long-term benefits of inhaled tobramycin in adolescent patients with cystic fibrosis. Chest (in press)
CHEST / 120 / 3 / SEPTEMBER, 2001 SUPPLEMENT
113S