Effect of Macrolides on In Vivo Ion Transport across Cystic Fibrosis Nasal Epithelium
http://www.100md.com
美国呼吸和危急护理医学 2005年第4期
Department of Pediatrics, University of North Carolina at Chapel Hill, Chapel Hill
Department of Pediatrics, Wake Forest University School of Medicine, Winston-Salem, North Carolina
Brown Medical School, Providence, Rhode Island
ABSTRACT
Fourteen- and 15-member macrolide antibiotics are under investigation as potential therapeutic agents for cystic fibrosis (CF). The nonantibiotic mechanisms of action of these compounds in CF are not understood. We used nasal potential difference (NPD) measurements to test the effect of macrolides on airway epithelial ion (chloride, sodium) transport of CF mice and humans. We tested clarithromycin and azithromycin in mice, and clarithromycin in patients with CF. Baseline and post-treatment NPD was measured in two strains (C57Bl6 and BalbC) of CF transmembrane regulator "knockout" and littermate control mice, and in F508/F508 mice. In addition, NPD was measured in 18 human subjects with CF (17 F-508/F-508 and 1 F-508/other) who were undergoing a 12-month, randomized, double-blind crossover study of the effects of clarithromycin on pulmonary outcome in CF. Neither clarithromycin nor azithromycin affected ion transport characteristics of normal or CF nasal epithelium in either mouse or humans. We conclude that the apparent beneficial effects of macrolides on pulmonary outcome in CF are not mediated by their modulation of ion transport.
Key Words: antibiotic chloride nasal potential difference therapy
Macrolide antibiotics are receiving increasing attention for their possible therapeutic benefits in the treatment of cystic fibrosis (CF). Recently, several preliminary and multicenter reports have supported the possibility that macrolides may improve lung function and other pulmonary outcomes in CF (1). The mechanism by which this effect occurs is not known. Pseudomonas aeruginosa shows resistance to macrolide antibiotics when grown under conventional methods, but it has been suggested that these antibiotics may inhibit P. aeruginosa by indirect action (2, 3). However, in addition to their possible antibacterial effects, macrolides have diverse nonantibiotic properties, including motilin receptor stimulation, interference with biofilm formation, and immunomodulatory effects (4). It has also been suggested that macrolides might exert their effect by normalizing ion transport across the CF respiratory epithelium (1, 5). In a preliminary report, 6 of 10 patients with a range of CF mutations showed a transient normalization of CleC permeability of the nasal epithelium after a 4-week course of azithromycin. In the present study, we used nasal potential difference (NPD) measurements to test the hypothesis that macrolides exert their beneficial effect by moderating the abnormal CF ion transport phenotype. In addition to human studies, the effect of macrolides on airway ion transport was also tested in mice, whose nasal epithelium is thought to provide an excellent model of human normal and CF airway ion transport. In mice, we tested whether macrolides might modulate Na+ hyperabsorption (CF transmembrane regulator [CFTR]eC/eC mice and F-508/F-508 mice), induce alternative CleC secretory pathways (CFTReC/eC mice), or perhaps increase CleC secretion via improved CFTR trafficking (F-508/F-508 mice). Some of the results of these studies have been reported previously in the form of an abstract (6).
METHODS
Mice
Two strains of CFTReC/eC (7) (BALB/C, n = 11; C57/BL6, n = 10) and one strain of CFTR mutant mice (8) (F508/F508, n = 4) were studied. Mouse age ranged from 6 to 16 weeks. Clarithromycin (2 mg/day suspension in water; Abbott Laboratories, Chicago, IL), azithromycin (2 mg/day suspension in water; Pfizer, New York, NY), or placebo (water) was delivered for 4 days by daily oral gavage after the baseline NPD had been measured. NPD was measured before and within 6 hours of the last clarithromycin dose. Blood for clarithromycin measurement was obtained after the second NPD.
Humans
Thirty patients with CF were enrolled in a randomized, double-blind, placebo-controlled, crossover study to evaluate the mechanism of effect of clarithromycin on the airway in patients with CF. Of these, 18 patients consented to participate in the NPD arm of the study (17 F508 homozygote and 1 F508/unknown mutation). Mean age was 14.8 years (range 6eC27 years). Patients were randomly assigned to receive 5 months of placebo or clarithromycin (Biaxin XL, 500 mg daily; Abbott Laboratories) and were then crossed over to the alternative treatment regimen for 5 months after a washout period (1 month). NPD was measured at entry to the study (n = 17), at the end of the first 5-month treatment period (n = 12), and at the end of the second 5-month treatment period (10). Nine subjects were evaluated at all three measurement points. NPDs were not obtained in some instances because of logistic issues (10) or patient illness (4). Data were excluded from one subject whose clarithromycin serum levels were below detection when he was in the active treatment period.
NPD
NPD was measured with the technique of Knowles and others (9) and modified for use in children, as described recently (10).
Mice were anesthetized with isoflurane (2eC3% vol/vol in O2) or ketamine/xylazine (4 ml/g body weight). The nasal cannula was inserted to a distance of 3 mm from the nasal orifice. A reference bridge was placed in the subcutaneous space of the hind leg. In mice, the perfusion rate was constant at 10 e/min. In humans, solutions 1 and 2 were perfused at 0.1 ml/min. Solutions 3, 4, and 5 were perfused at 1.2 ml/min to ensure flooding of the nasal surface with CleC free solutions. Five solutions were perfused in succession: Lactated Ringers, Ringers with amiloride (10eC4 M) (Amil), CleC free Krebs Bicarbonate Ringers (containing 10eC4 M amiloride) (CleC free), CleC free Krebs Bicarbonate Ringers with 10eC5 M isoproterenol (Iso), CleC free with isoproterenol, and 10eC4 M adenosine triphosphate.
Clarithromycin Levels
For mice, 0.2 ml blood was obtained at the termination of the study (4 hours after the last dose) under anesthesia by cardiac puncture. For human studies, blood was obtained at the time of NPD measurement. Clarithromycin was measured by bioassay as published previously (11).
Statistics
Differences in NPD parameters before and after intervention (pretreatment to treatment or placebo) for both study periods (before crossover and after crossover) were tested by paired t tests. For human studies, we included the results from the subjects who received clarithromycin before crossover with those who had received clarithromycin after crossover because the latter group had received placebo only in the first half of the study. Power calculations were performed using the xsampsi procedure in STATA7 (Stata Corporation, College Station, TX) (12). For mouse studies, power calculations were performed using the sampsi procedure in STADA7.
RESULTS
Macrolide Blood Levels
Blood levels for mice dosed with 2 mg clarithromycin per day were 1.0 ± 0.3 e/ml, and levels for mice dosed with placebo were all below the detection limit of 0.050 e/ml. In human subjects, adherence to the treatment schedule as determined by pill counting correlated with the presence of clarithromycin in serum. One patient who had been prescribed clarithromycin had undetectable levels of the drug in his serum and did not present for his NPD at the end of the treatment period. Of the remaining subjects who had been prescribed macrolides, mean serum clarithromycin level at the time of NPD study was 1,236 ng/ml (range 23eC2700 ng/ml).
NPD
Mouse studies.
CFTReC/eC and F-508/F-508 mice had the anticipated bioelectric profile of CF mouse nasal epithelium (i.e., high basal PD, increased amiloride sensitive PD, absent hyperpolarization of NPD after perfusion with zero CleC solution, and increased hyperpolarization after adenosine triphosphate compared with littermate controls) (Figure 1). After 4 days of clarithromycin treatment, the bioelectric values were unchanged in placebo- and macrolide-treated mice (Table 1). Because of the previous report suggesting increased CleC permeability in human CF nasal epithelium after azithromycin treatment, six C57Bl6 CFTReC/eC mice were dosed with azithromycin 2 mg/day (3) or placebo (3) for 4 days. There were no changes in the typical CF bioelectric profiles measured at the start and at the end of the dosing regime (Table 1). We also tested the hypothesis that macrolides worked by increasing trafficking of defective CFTR to the membrane. For these experiments, we treated F508 mice with clarithromycin. Again, we found no effect of clarithromycin on nasal ion transport (Table 1).
When all placebo-treated mice were compared with all macrolide-treated mice, the power of this study to detect a clinically significant change in NPD after perfusion of CleC free solution containing isoproterenol (PDCl FREE) of ( eC5 mV) was 98%. In subgroup analysis, power to detect similar change was as follows: placebo versus clarithromycin, 87.8%; azithromycin, 86.6%.
Human studies.
PD values obtained at the start of the study were typical of those reported for patients with CF (9). We observed no changes in NPD profile when subjects received clarithromycin or placebo (Figure 2). Although the number of patients included in this study was small, the power of this study to detect a clinically significant change in PDCl FREE greater than or equal to 10 mV was 79.1%.
DISCUSSION
Randomized controlled trials evaluating macrolides in the treatment of CF lung disease have shown a 3.5 to 5.5% improvement in FEV1 (13, 14). Preliminary reports of a larger randomized controlled trials with 185 participants suggest that sustained improvement in lung function occurs in patients with CF who take chronic macrolide therapy (15).
Macrolides kill a wide range of bacteria that cause respiratory disease, but are only modestly active against P. aeruginosa or Staphylococcus aureus, the two primary airway pathogens in CF. Interest in the use of macrolides stems from their apparent effectiveness in the treatment of diffuse panbronchiolitis, an inflammatory airway disease similar to CF (16). Diffuse panbronchiolitis, the most commonly reported cause of bronchiectasis in East Asia, usually begins in childhood with persistent sinusitis but can progress to P. aeruginosa infection (17). Previous reports suggest that the long-term use of macrolides has a beneficial affect on outcome of these patients (16). This has been attributed to direct antiinflammatory properties of this class of antibiotic (18eC20).
Despite the in vitro evidence of anti-inflammatory and antivirulence effects of macrolides, the precise mechanism of these apparent beneficial effects on CF lung disease is not known. Deranged respiratory epithelial ion transport is thought to be a major contributor to lung disease in CF. However, the possibility that macrolides may improve CleC secretion in CF was raised in a report of a small group (10) of patients with CF who were treated with daily azithromycin for 1 month (5). In 6 of 10 patients, there was normalization of nasal epithelial chloride secretion after azithromycin therapy, as evidenced by hyperpolarization of NPD after perfusion of a chloride-free solution. In contrast to the response in normal subjects, this hyperpolarization was not sustained after subsequent perfusion with isoproterenol. In that study, azithromycin did not significantly affect the other hallmark ion transport defects in CF such as elevation of the basal potential difference, and the large component of basal PD inhibited by amiloride. This CleC transport modulatory effect of macrolides on ion transport has not been reported previously. In fact, previous In vitro and in vivo studies suggested that macrolides may inhibit both CleC secretion and Na+ absorption across sheep tracheal epithelial cells (21) and human nasal epithelium (22), respectively.
In contrast to the study of Pradal and others (5), our human studies were randomized, double-blinded, and placebo-controlled, and our intervention was given over a longer period (5 months versus 4 weeks). The design in our study included a crossover arm and a 1-month washout period. NPD measurements were spaced far enough apart to minimize any carryover effect of the active drug after crossover. Pradal and coworkers reported NPD findings from the 6 of 10 subjects who had shown an apparent CleC secretory response (5). When our entire group of subjects were analyzed, no NPD parameters (basal PD, amiloride-inhibitable PD, NPD response to CleC free solution) were affected by clarithromycin therapy or by withdrawal of clarithromycin therapy. It is possible that some of our patients could have received variable amount of drug during the 5-month period, but pill counting suggests that noncompliance with medications was not a significant issue. Although some NPD parameters moved toward the normal range in some clarithromycin-treated patients, this is attributable to random fluctuation of NPD. No clarithromycin-treated patient demonstrated normalization of NPD parameters with subsequent regression to more abnormal results after switching to placebo. Although the number of human subjects was small, by using a crossover research design, we were able to attain adequate power to ensure that the absence of drug effect is unlikely to be attributable to a Type II () error.
It is possible that some of the differences effect of macrolide observed between our study and that of Pradal and others (5) could be due to differences in effect of clarithromycin and azithromycin on ion transport. Mouse studies provided an opportunity to test this possible effect of drug type. In CF knockout mice, neither clarithromycin nor azithromycin affected ion transport. We studied both CFTR "null" mice (CFTReC/eC) and F-508 mutant mice (CFTR mut/mut), which are more representative of the human subjects who participated in our clinical trial. Treatment of DF508 mice with clarithromycin resulted in a small change in Cl permeability that was not statistically significant. Although the small numbers of mice used may have masked a statistical difference, the magnitude of the change we observed (eC2.7 mV) is unlikely to be clinically significant.
The absence of a significant change in abnormal patterns of nasal Na+ or CleC transport in either human or murine studies supports our conclusion that the beneficial effect of macrolide antibiotics in patients with CF is not mediated by amelioration of the ion transport defects across respiratory epithelia.
Acknowledgments
The authors thank Lauren Clarkson for assistance in sample collection and Amy Cook, Pharm D, Abbott Laboratories, for assistance with measurement of clarithromycin levels in mice.
REFERENCES
Bush A, Rubin BK. Macrolides as biologic response modifiers in cystic fibrosis and bronchiectasis. Semin Resp Crit Care Med 2003;24:737eC748.
Tateda K, Hirakata Y, Furuya N, Ohno A, Yamaguchi K. Effects of sub-MICs of erythromycin and other macrolide antibiotics on serum sensitivity of Pseudomonas aeruginosa. Antimicrob Agents Chemother 1993;37:675eC680.
Tateda K, Ishii Y, Matsumoto T, Furuya N, Nagashima M, Matsunaga T, Ohno A, Miyazaki S, Yamaguchi K. Direct evidence for antipseudomonal activity of macrolides: exposure-dependent bactericidal activity and inhibition of protein synthesis by erythromycin, clarithromycin, and azithromycin. Antimicrob Agents Chemother 1996;40:2271eC2275.
Rubin BK, Tamaoki J. Macrolide antibiotics as biological response modifiers. Curr Opin Investig Drugs 2000;1:169eC172.
Pradal U, Delmarco A, Cipoli M, Cazzoa G. Chloride transport may be restored by long-term azithromycin treatment in patients with cystic fibrosis. Pediatr Pulmonol 2000;(Suppl20):280.
Gillie DJ, Barker PM. Effect of clarithromycin on in vivo ion transport by CFTR eC/eC mouse nasal epithelium. Pediatr Pulmonol 2001;(Suppl22):259.
Snouwaert JN, Brigman KK, Latour AM, Malouf NN, Boucher RC, Smithies O, Koller BH. An animal model for cystic fibrosis made by gene targeting. Science 1992;257:1083eC1088.
Zeiher BG, Eichwald E, Zabner J, Smith JJ, Puga AP, McCray PB Jr, Capecchi MR, Welsh MJ, Thomas KR. A mouse model for the delta F508 allele of cystic fibrosis. J Clin Invest 1995;96:2051eC2064.
Knowles MR, Paradiso AM, Boucher RC. In vivo nasal potential difference: techniques and protocols for assessing efficacy of gene transfer in cystic fibrosis. Hum Gene Ther 1995;6:445eC455.
Southern KW, Noone PG, Bosworth DG, Legrys VA, Knowles MR, Barker PM. A modified technique for measurement of nasal transepithelial potential difference in infants. J Pediatr 2001;139:353eC358.
Fernandes PB, Ramer N, Rode RA, Freiberg L. Bioassay for A-56268 (TE-031) and identification of its major metabolite, 14-hydroxy-6-O-methyl erythromycin. Eur J Clin Microbiol Infect Dis 1988;7:73eC76.
Senn S. Crossover trials in clinical research, 2nd ed. New York: John Wiley & Sons; 2002.
Wolter J, Seeney S, Bell S, Bowler S, Masel P, McCormack J. Effect of long term treatment with azithromycin on disease parameters in cystic fibrosis: a randomised trial. Thorax 2002;57:212eC216.
Equi A, Balfour-Lynn IM, Bush A, Rosenthal M. Long term azithromycin in children with cystic fibrosis: a randomised, placebo-controlled crossover trial. Lancet 2002;360:978eC984.
Saiman L, Marshall BC, Mayer-Hamblett N, Burns JL, Quittner AL, Cibene DA, Coquillette S, Fieberg AY, Accurso FJ, Campbell PW III, et al. Azithromycin in patients with cystic fibrosis chronically infected with Pseudomonas aeruginosa: a randomized controlled trial. JAMA 2003;290:1749eC1756.
Kobayashi H. Biofilm disease: its clinical manifestation and therapeutic possibilities of macrolides. Am J Med 1995;99:26SeC30S.
Hoiby N. Diffuse panbronchiolitis and cystic fibrosis: East meets West. Thorax 1994;49:531eC532.
Labro MT. Anti-inflammatory activity of macrolides: a new therapeutic potential J Antimicrob Chemother 1998;41(Suppl B):37eC46.
Anderson R, Theron AJ, Feldman C. Membrane-stabilizing, anti-inflammatory interactions of macrolides with human neutrophils. Inflammation 1996;20:693eC705.
Yanagihara K, Tomono K, Sawai T, Hirakata Y, Kadota J, Koga H, Tashiro T, Kohno S. Effect of clarithromycin on lymphocytes in chronic respiratory Pseudomonas aeruginosa infection. Am J Respir Crit Care Med 1997;155:337eC342.
Tamaoki J, Isono K, Sakai N, Kanemura T, Konno K. Erythromycin inhibits Cl secretion across canine tracheal epithelial cells. Eur Respir J 1992;5:234eC238.
Middleton PG, Geddes DM, Alton EW. Trimethoprim and tetracycline inhibit airway epithelial sodium absorption. Am J Respir Crit Care Med 1996;154:18eC23.(Pierre M. Barker, Daniel )
Department of Pediatrics, Wake Forest University School of Medicine, Winston-Salem, North Carolina
Brown Medical School, Providence, Rhode Island
ABSTRACT
Fourteen- and 15-member macrolide antibiotics are under investigation as potential therapeutic agents for cystic fibrosis (CF). The nonantibiotic mechanisms of action of these compounds in CF are not understood. We used nasal potential difference (NPD) measurements to test the effect of macrolides on airway epithelial ion (chloride, sodium) transport of CF mice and humans. We tested clarithromycin and azithromycin in mice, and clarithromycin in patients with CF. Baseline and post-treatment NPD was measured in two strains (C57Bl6 and BalbC) of CF transmembrane regulator "knockout" and littermate control mice, and in F508/F508 mice. In addition, NPD was measured in 18 human subjects with CF (17 F-508/F-508 and 1 F-508/other) who were undergoing a 12-month, randomized, double-blind crossover study of the effects of clarithromycin on pulmonary outcome in CF. Neither clarithromycin nor azithromycin affected ion transport characteristics of normal or CF nasal epithelium in either mouse or humans. We conclude that the apparent beneficial effects of macrolides on pulmonary outcome in CF are not mediated by their modulation of ion transport.
Key Words: antibiotic chloride nasal potential difference therapy
Macrolide antibiotics are receiving increasing attention for their possible therapeutic benefits in the treatment of cystic fibrosis (CF). Recently, several preliminary and multicenter reports have supported the possibility that macrolides may improve lung function and other pulmonary outcomes in CF (1). The mechanism by which this effect occurs is not known. Pseudomonas aeruginosa shows resistance to macrolide antibiotics when grown under conventional methods, but it has been suggested that these antibiotics may inhibit P. aeruginosa by indirect action (2, 3). However, in addition to their possible antibacterial effects, macrolides have diverse nonantibiotic properties, including motilin receptor stimulation, interference with biofilm formation, and immunomodulatory effects (4). It has also been suggested that macrolides might exert their effect by normalizing ion transport across the CF respiratory epithelium (1, 5). In a preliminary report, 6 of 10 patients with a range of CF mutations showed a transient normalization of CleC permeability of the nasal epithelium after a 4-week course of azithromycin. In the present study, we used nasal potential difference (NPD) measurements to test the hypothesis that macrolides exert their beneficial effect by moderating the abnormal CF ion transport phenotype. In addition to human studies, the effect of macrolides on airway ion transport was also tested in mice, whose nasal epithelium is thought to provide an excellent model of human normal and CF airway ion transport. In mice, we tested whether macrolides might modulate Na+ hyperabsorption (CF transmembrane regulator [CFTR]eC/eC mice and F-508/F-508 mice), induce alternative CleC secretory pathways (CFTReC/eC mice), or perhaps increase CleC secretion via improved CFTR trafficking (F-508/F-508 mice). Some of the results of these studies have been reported previously in the form of an abstract (6).
METHODS
Mice
Two strains of CFTReC/eC (7) (BALB/C, n = 11; C57/BL6, n = 10) and one strain of CFTR mutant mice (8) (F508/F508, n = 4) were studied. Mouse age ranged from 6 to 16 weeks. Clarithromycin (2 mg/day suspension in water; Abbott Laboratories, Chicago, IL), azithromycin (2 mg/day suspension in water; Pfizer, New York, NY), or placebo (water) was delivered for 4 days by daily oral gavage after the baseline NPD had been measured. NPD was measured before and within 6 hours of the last clarithromycin dose. Blood for clarithromycin measurement was obtained after the second NPD.
Humans
Thirty patients with CF were enrolled in a randomized, double-blind, placebo-controlled, crossover study to evaluate the mechanism of effect of clarithromycin on the airway in patients with CF. Of these, 18 patients consented to participate in the NPD arm of the study (17 F508 homozygote and 1 F508/unknown mutation). Mean age was 14.8 years (range 6eC27 years). Patients were randomly assigned to receive 5 months of placebo or clarithromycin (Biaxin XL, 500 mg daily; Abbott Laboratories) and were then crossed over to the alternative treatment regimen for 5 months after a washout period (1 month). NPD was measured at entry to the study (n = 17), at the end of the first 5-month treatment period (n = 12), and at the end of the second 5-month treatment period (10). Nine subjects were evaluated at all three measurement points. NPDs were not obtained in some instances because of logistic issues (10) or patient illness (4). Data were excluded from one subject whose clarithromycin serum levels were below detection when he was in the active treatment period.
NPD
NPD was measured with the technique of Knowles and others (9) and modified for use in children, as described recently (10).
Mice were anesthetized with isoflurane (2eC3% vol/vol in O2) or ketamine/xylazine (4 ml/g body weight). The nasal cannula was inserted to a distance of 3 mm from the nasal orifice. A reference bridge was placed in the subcutaneous space of the hind leg. In mice, the perfusion rate was constant at 10 e/min. In humans, solutions 1 and 2 were perfused at 0.1 ml/min. Solutions 3, 4, and 5 were perfused at 1.2 ml/min to ensure flooding of the nasal surface with CleC free solutions. Five solutions were perfused in succession: Lactated Ringers, Ringers with amiloride (10eC4 M) (Amil), CleC free Krebs Bicarbonate Ringers (containing 10eC4 M amiloride) (CleC free), CleC free Krebs Bicarbonate Ringers with 10eC5 M isoproterenol (Iso), CleC free with isoproterenol, and 10eC4 M adenosine triphosphate.
Clarithromycin Levels
For mice, 0.2 ml blood was obtained at the termination of the study (4 hours after the last dose) under anesthesia by cardiac puncture. For human studies, blood was obtained at the time of NPD measurement. Clarithromycin was measured by bioassay as published previously (11).
Statistics
Differences in NPD parameters before and after intervention (pretreatment to treatment or placebo) for both study periods (before crossover and after crossover) were tested by paired t tests. For human studies, we included the results from the subjects who received clarithromycin before crossover with those who had received clarithromycin after crossover because the latter group had received placebo only in the first half of the study. Power calculations were performed using the xsampsi procedure in STATA7 (Stata Corporation, College Station, TX) (12). For mouse studies, power calculations were performed using the sampsi procedure in STADA7.
RESULTS
Macrolide Blood Levels
Blood levels for mice dosed with 2 mg clarithromycin per day were 1.0 ± 0.3 e/ml, and levels for mice dosed with placebo were all below the detection limit of 0.050 e/ml. In human subjects, adherence to the treatment schedule as determined by pill counting correlated with the presence of clarithromycin in serum. One patient who had been prescribed clarithromycin had undetectable levels of the drug in his serum and did not present for his NPD at the end of the treatment period. Of the remaining subjects who had been prescribed macrolides, mean serum clarithromycin level at the time of NPD study was 1,236 ng/ml (range 23eC2700 ng/ml).
NPD
Mouse studies.
CFTReC/eC and F-508/F-508 mice had the anticipated bioelectric profile of CF mouse nasal epithelium (i.e., high basal PD, increased amiloride sensitive PD, absent hyperpolarization of NPD after perfusion with zero CleC solution, and increased hyperpolarization after adenosine triphosphate compared with littermate controls) (Figure 1). After 4 days of clarithromycin treatment, the bioelectric values were unchanged in placebo- and macrolide-treated mice (Table 1). Because of the previous report suggesting increased CleC permeability in human CF nasal epithelium after azithromycin treatment, six C57Bl6 CFTReC/eC mice were dosed with azithromycin 2 mg/day (3) or placebo (3) for 4 days. There were no changes in the typical CF bioelectric profiles measured at the start and at the end of the dosing regime (Table 1). We also tested the hypothesis that macrolides worked by increasing trafficking of defective CFTR to the membrane. For these experiments, we treated F508 mice with clarithromycin. Again, we found no effect of clarithromycin on nasal ion transport (Table 1).
When all placebo-treated mice were compared with all macrolide-treated mice, the power of this study to detect a clinically significant change in NPD after perfusion of CleC free solution containing isoproterenol (PDCl FREE) of ( eC5 mV) was 98%. In subgroup analysis, power to detect similar change was as follows: placebo versus clarithromycin, 87.8%; azithromycin, 86.6%.
Human studies.
PD values obtained at the start of the study were typical of those reported for patients with CF (9). We observed no changes in NPD profile when subjects received clarithromycin or placebo (Figure 2). Although the number of patients included in this study was small, the power of this study to detect a clinically significant change in PDCl FREE greater than or equal to 10 mV was 79.1%.
DISCUSSION
Randomized controlled trials evaluating macrolides in the treatment of CF lung disease have shown a 3.5 to 5.5% improvement in FEV1 (13, 14). Preliminary reports of a larger randomized controlled trials with 185 participants suggest that sustained improvement in lung function occurs in patients with CF who take chronic macrolide therapy (15).
Macrolides kill a wide range of bacteria that cause respiratory disease, but are only modestly active against P. aeruginosa or Staphylococcus aureus, the two primary airway pathogens in CF. Interest in the use of macrolides stems from their apparent effectiveness in the treatment of diffuse panbronchiolitis, an inflammatory airway disease similar to CF (16). Diffuse panbronchiolitis, the most commonly reported cause of bronchiectasis in East Asia, usually begins in childhood with persistent sinusitis but can progress to P. aeruginosa infection (17). Previous reports suggest that the long-term use of macrolides has a beneficial affect on outcome of these patients (16). This has been attributed to direct antiinflammatory properties of this class of antibiotic (18eC20).
Despite the in vitro evidence of anti-inflammatory and antivirulence effects of macrolides, the precise mechanism of these apparent beneficial effects on CF lung disease is not known. Deranged respiratory epithelial ion transport is thought to be a major contributor to lung disease in CF. However, the possibility that macrolides may improve CleC secretion in CF was raised in a report of a small group (10) of patients with CF who were treated with daily azithromycin for 1 month (5). In 6 of 10 patients, there was normalization of nasal epithelial chloride secretion after azithromycin therapy, as evidenced by hyperpolarization of NPD after perfusion of a chloride-free solution. In contrast to the response in normal subjects, this hyperpolarization was not sustained after subsequent perfusion with isoproterenol. In that study, azithromycin did not significantly affect the other hallmark ion transport defects in CF such as elevation of the basal potential difference, and the large component of basal PD inhibited by amiloride. This CleC transport modulatory effect of macrolides on ion transport has not been reported previously. In fact, previous In vitro and in vivo studies suggested that macrolides may inhibit both CleC secretion and Na+ absorption across sheep tracheal epithelial cells (21) and human nasal epithelium (22), respectively.
In contrast to the study of Pradal and others (5), our human studies were randomized, double-blinded, and placebo-controlled, and our intervention was given over a longer period (5 months versus 4 weeks). The design in our study included a crossover arm and a 1-month washout period. NPD measurements were spaced far enough apart to minimize any carryover effect of the active drug after crossover. Pradal and coworkers reported NPD findings from the 6 of 10 subjects who had shown an apparent CleC secretory response (5). When our entire group of subjects were analyzed, no NPD parameters (basal PD, amiloride-inhibitable PD, NPD response to CleC free solution) were affected by clarithromycin therapy or by withdrawal of clarithromycin therapy. It is possible that some of our patients could have received variable amount of drug during the 5-month period, but pill counting suggests that noncompliance with medications was not a significant issue. Although some NPD parameters moved toward the normal range in some clarithromycin-treated patients, this is attributable to random fluctuation of NPD. No clarithromycin-treated patient demonstrated normalization of NPD parameters with subsequent regression to more abnormal results after switching to placebo. Although the number of human subjects was small, by using a crossover research design, we were able to attain adequate power to ensure that the absence of drug effect is unlikely to be attributable to a Type II () error.
It is possible that some of the differences effect of macrolide observed between our study and that of Pradal and others (5) could be due to differences in effect of clarithromycin and azithromycin on ion transport. Mouse studies provided an opportunity to test this possible effect of drug type. In CF knockout mice, neither clarithromycin nor azithromycin affected ion transport. We studied both CFTR "null" mice (CFTReC/eC) and F-508 mutant mice (CFTR mut/mut), which are more representative of the human subjects who participated in our clinical trial. Treatment of DF508 mice with clarithromycin resulted in a small change in Cl permeability that was not statistically significant. Although the small numbers of mice used may have masked a statistical difference, the magnitude of the change we observed (eC2.7 mV) is unlikely to be clinically significant.
The absence of a significant change in abnormal patterns of nasal Na+ or CleC transport in either human or murine studies supports our conclusion that the beneficial effect of macrolide antibiotics in patients with CF is not mediated by amelioration of the ion transport defects across respiratory epithelia.
Acknowledgments
The authors thank Lauren Clarkson for assistance in sample collection and Amy Cook, Pharm D, Abbott Laboratories, for assistance with measurement of clarithromycin levels in mice.
REFERENCES
Bush A, Rubin BK. Macrolides as biologic response modifiers in cystic fibrosis and bronchiectasis. Semin Resp Crit Care Med 2003;24:737eC748.
Tateda K, Hirakata Y, Furuya N, Ohno A, Yamaguchi K. Effects of sub-MICs of erythromycin and other macrolide antibiotics on serum sensitivity of Pseudomonas aeruginosa. Antimicrob Agents Chemother 1993;37:675eC680.
Tateda K, Ishii Y, Matsumoto T, Furuya N, Nagashima M, Matsunaga T, Ohno A, Miyazaki S, Yamaguchi K. Direct evidence for antipseudomonal activity of macrolides: exposure-dependent bactericidal activity and inhibition of protein synthesis by erythromycin, clarithromycin, and azithromycin. Antimicrob Agents Chemother 1996;40:2271eC2275.
Rubin BK, Tamaoki J. Macrolide antibiotics as biological response modifiers. Curr Opin Investig Drugs 2000;1:169eC172.
Pradal U, Delmarco A, Cipoli M, Cazzoa G. Chloride transport may be restored by long-term azithromycin treatment in patients with cystic fibrosis. Pediatr Pulmonol 2000;(Suppl20):280.
Gillie DJ, Barker PM. Effect of clarithromycin on in vivo ion transport by CFTR eC/eC mouse nasal epithelium. Pediatr Pulmonol 2001;(Suppl22):259.
Snouwaert JN, Brigman KK, Latour AM, Malouf NN, Boucher RC, Smithies O, Koller BH. An animal model for cystic fibrosis made by gene targeting. Science 1992;257:1083eC1088.
Zeiher BG, Eichwald E, Zabner J, Smith JJ, Puga AP, McCray PB Jr, Capecchi MR, Welsh MJ, Thomas KR. A mouse model for the delta F508 allele of cystic fibrosis. J Clin Invest 1995;96:2051eC2064.
Knowles MR, Paradiso AM, Boucher RC. In vivo nasal potential difference: techniques and protocols for assessing efficacy of gene transfer in cystic fibrosis. Hum Gene Ther 1995;6:445eC455.
Southern KW, Noone PG, Bosworth DG, Legrys VA, Knowles MR, Barker PM. A modified technique for measurement of nasal transepithelial potential difference in infants. J Pediatr 2001;139:353eC358.
Fernandes PB, Ramer N, Rode RA, Freiberg L. Bioassay for A-56268 (TE-031) and identification of its major metabolite, 14-hydroxy-6-O-methyl erythromycin. Eur J Clin Microbiol Infect Dis 1988;7:73eC76.
Senn S. Crossover trials in clinical research, 2nd ed. New York: John Wiley & Sons; 2002.
Wolter J, Seeney S, Bell S, Bowler S, Masel P, McCormack J. Effect of long term treatment with azithromycin on disease parameters in cystic fibrosis: a randomised trial. Thorax 2002;57:212eC216.
Equi A, Balfour-Lynn IM, Bush A, Rosenthal M. Long term azithromycin in children with cystic fibrosis: a randomised, placebo-controlled crossover trial. Lancet 2002;360:978eC984.
Saiman L, Marshall BC, Mayer-Hamblett N, Burns JL, Quittner AL, Cibene DA, Coquillette S, Fieberg AY, Accurso FJ, Campbell PW III, et al. Azithromycin in patients with cystic fibrosis chronically infected with Pseudomonas aeruginosa: a randomized controlled trial. JAMA 2003;290:1749eC1756.
Kobayashi H. Biofilm disease: its clinical manifestation and therapeutic possibilities of macrolides. Am J Med 1995;99:26SeC30S.
Hoiby N. Diffuse panbronchiolitis and cystic fibrosis: East meets West. Thorax 1994;49:531eC532.
Labro MT. Anti-inflammatory activity of macrolides: a new therapeutic potential J Antimicrob Chemother 1998;41(Suppl B):37eC46.
Anderson R, Theron AJ, Feldman C. Membrane-stabilizing, anti-inflammatory interactions of macrolides with human neutrophils. Inflammation 1996;20:693eC705.
Yanagihara K, Tomono K, Sawai T, Hirakata Y, Kadota J, Koga H, Tashiro T, Kohno S. Effect of clarithromycin on lymphocytes in chronic respiratory Pseudomonas aeruginosa infection. Am J Respir Crit Care Med 1997;155:337eC342.
Tamaoki J, Isono K, Sakai N, Kanemura T, Konno K. Erythromycin inhibits Cl secretion across canine tracheal epithelial cells. Eur Respir J 1992;5:234eC238.
Middleton PG, Geddes DM, Alton EW. Trimethoprim and tetracycline inhibit airway epithelial sodium absorption. Am J Respir Crit Care Med 1996;154:18eC23.(Pierre M. Barker, Daniel )