Acute and Long-term Amiloride Inhalation in Cystic Fibrosis Lung Disease A Rational Approach to Cystic Fibrosis Therapy1-3
ERNST M. APP, MALCOLM KING, RENATE HELFESRIEDER, DIETER KOHLER, and HEINRICH MATTHYS
Introduction Cystic fibrosis (CF), first described just 50 yr ago, is today the most common lethal inherited disease among Caucasians (1).This autosomal recessive disorder has a carrier rate ranging from approximately 1:20 in most of Europe and among white Americans to 1:60 in U.S. blacks and 1:150in the Japanese of Hawaii. The basic defect relates to regulation of ion transport by the secretory epithelia. A reduced chloride conductance occurs in both airway and sweat ductal epithelia in CF (2). In addition, in airway epithelium, but not in sweat gland, the sodium absorption is abnormally elevated (2, 3). Thus, if augmented volume absorption and/or reduced water secretion playa role in the pathogenesis of CF airway disease, then modalities to either induce CI- secretion or block the excess Na' absorption may be beneficial. Because no established drug activates the defective CI- channels in CF airway epithelia, the alternative concept, i.e., blocking Na" absorption, appears to be the most rational approach to develop a useful therapy in CF. Investigations have shown that bronchial clearance is impaired in patients with CF (4). This may be due to rheologic abnormalities in the CF mucus, which could arise from dysfunction of transepithelial electrolyte movement (5). In CF, the respiratory secretions exhibit decreased sodium and chloride contents (6), only partially compensated by increased potassium and other electrolytes, which appear to reflect the altered ion transport mechanisms. Whether this results in a decreased water content and hence increased viscosity or rigidity is not yet certain, although the water content of CF sputum (~ 89010) has been reported to be lower than in other lung diseases (~ 95010) (7) or than that of normal canine tracheal mucus (~ 93070) (8, 9). Thus, alterations in electrolyte content might corre-
SUMMARY Cystic fibrosis (CF) Is the most common Inherited fatal disorder among caucasians. Bronchial mucus In CF contains more potassium and less sodium, which may be due to Increased sodium absorption, resulting In a reduced airway water content. We studied 23 Patients with CF after Inhalation of normal saline or amllorlde (10-3 M), a sodium transport bloc.r. Mucocillary clearance (MC) and cough clearance (CC) were determined with a gamma camera that traced the movement of """fc-Iabeled, hardened erythrocytes over a 1-h Period after the patients Inhaled these particles as an aerosol. Before and after each Investigation pulmonary function tests (PFT) and blood pressure (BP) were measured. Sputum thread formation was measured by means of a fllancemeter. Six of the patients also completed a 3-wk trial of amllorlde Inhalation therapy. MC Increased significantly (p < 0.001) after acute amllorlde Inhalation (bronchial deposition, O.fIT mg amllorlde) compared with that In the saline control. CC also Increased, but not as much as MC.After 3 wk of amlloride Inhalation (2 times a day) clearance values (both MC and CC) were markedly enhanced (p < 0.01); after a similar Period of saline Inhalation, clearance values were not different from baseline. Sputum fllance values also decreased significantly after amllorlde Inhalation. There were no adverse effects of the amllorlde Inhalation compared with saline. We conclude that amllorlde Inhal. tlon administered as a single dose or as long-term therapy Is able to Incre_ MC and CC In CF airways and that the effect of 10-3 M amllorlde Inhalation on MC lasts at least 40 min. These results Indicate that further clinical studies of long-term amllorlde Inhalation In Patients with CF should be undertaken. This Is the first line of evidence that an Inhaled diuretic drug such as amllorlde Is able to enhance mucus clearance favorably In patients with CF. The mechanism Is most lliely due to the alteration of water content within the airways. AM REV RESPIR DIS 1990; 141:605-612
late with alterations in sputum rheology, which in tum might relate to impairments in ciliary and/or cough clearability of mucus. These suggestions led us to study amiloride (pyrazinecarbonylguanidine) as a novel therapeutic approach to this disease. Boucher and coworkers (3) and Knowles and colleagues (10) had previously shown that amiloride applied directly to the luminal side of the respiratory epithelium is able to "normalize" the increased transepithelial potential difference in patients with CF, and they proposed its use as a therapy in CF lung disease. Amiloride as a diuretic drug acts on the nephron in the kidney by blocking sodium reabsorption while preventing potassium excretion. For this to occur the amiloride must first be filtered in the glomerulus and act on the luminal side of the epithelium (11). A similar mechanism could result from the inhalation of amiloride with a reduction of the transepithelial potential difference and an in-
crease in the amount of sodium in pulmonary secretions. Curtis and Misch (12) showed a decreased mucociliary transport on the frog palate in vitro after amiloride, and they suggested this may be due to increased mucus discharge and altered mucus rheology. However, in a preliminary report (13), we were able to show an increase
(Received in original form April 10, 1989 and in revised form August 16, 1989) 1 From the Pulmonary Defense Group, University of Alberta, Edmonton, Alberta, Canada; the Division of Pulmonary Disease, Department of Internal Medicine, University of Freiburg, Freiburg i.Brsg., and the Fachkrankenhaus Kloster Grafschaft, Schmallenbrg, Federal Republic of Germany. 2 Supported by the Ministry for Science and Art of the Federal State of Baden-Wiirttemberg and by the Canadian Cystic Fibrosis Foundation. 3 Correspondence and requests for reprints should be addressed to Ernst M. App, M.D., Pulmonary Defense Group, 519 Newton Research Bldg, University of Alberta, Edmonton, Alberta TOO 2C2, Canada.
605
AP~ KING, HELFESRIEDER, KOHLER, AND MATTHYS
606
in mucus clearance as an acute effect of amiloride in CR Recently, Benedetto and coworkers (14)presented their results on cilia beat frequency in vitro and found a small but statistically significant increase in cilia beat frequency (CBF) with amiloride for approximately 1 h, which may account for the increased mucociliary clearance (MC) in CF that we previously demonstrated. We therefore wanted to examine the mechanisms for the acute improvement of MC, as well as cough clearance (CC), in CF after inhalation of amiloride. Furthermore, wewanted to explore the benefits of longer term or chronic amiloride inhalation on bronchial mucus clearance in such patients. Methods Patients I\venty-three patients (16 male and seven female) participated in the acute trial of amiloride therapy, including 14 patients described in the previous report (13).They were recruited from the outpatient clinic in Freiburg and from three other cystic fibrosis centers in the southwest of Germany. Their mean age (± SD) was 15.9 (± 6.2) with a range of 7 to 27 yr. Their height varied between 119and 181 em (mean, 153.9 ± 20.4 em) and their weight between 18and 65 kg (mean, 39.8 ± 13.4kg). They represented the whole spectrum from mild to severe symptomatic, from primarily gastrointestinal symptoms to mainly pulmonary, although no patients with acute exacerbations were studied. Of these 23 patients, six (three male and three female) subsequently underwent a 3-wk twice-daily inhalation trial with 10-3 M amiloride and the same period for placebo (0.9070 NaCI). The mean age (± SO) in this group was 17.1 (± 4.9) with a range of 12 to 23 yr. Their body height was between 139 and 179 em with an average of 161.3 (± 14.9), and a weight range between 23 and 54 kg with a mean of 41.3 (± 11.5). On the basis of their experience with the acute trial, the patients decided for themselves if they wanted to undertake this long-term therapy trial and were then entered at random in the study design. A possible bias may be that only "responder" patients decided to participate, but this did not turn out to be the case (see below). For each investigation the patient and/or the accompanying parent was informed about the procedure and the possible risks involved, and the consent was given in writing. The study was approved by the institutional ethics committee. Inclusion criteria. Confirmed diagnosis of CF. The diagnosis of CF in our patients was confirmed by typical history, clinical, radiographic, and sweat test findings. Exclusion criteria. (1) Pregnancy. (2) Patients with an acute pulmonary infection. (3) Patients younger than 7 yr of age, because
cooperation is required for aerosol inhalation and follow-up studies. Therapy changes. For the measurement of the acute effect of amiloride, we ensured that the patients did not alter their ongoing therapy between the two occasions of mucus clearance examination, which occurred on two separate days (at the same time of day) not longer than 1 wk apart. For the study of the chronic effect of amiloride, the patients were also maintained on their usual therapy (chest physiotherapy, inhaled sympathomimetics) except for mucolytic agents, which were withdrawn and replaced with amiloride or saline. None of the patients was receiving oral antibiotics during the chronic trial.
Drug Background and Concentration A dose-dependent effect of amiloride on airway bioelectric properties has been described over the concentration range of 10-7 to 10-3 M (12). Higher concentrations of drug, on the order of 10-2 M, are not soluble in physiological saline without reducing the overall electrolyte concentration. For the acute aerosol measurements, we chose to nebulize freshly prepared 10-3 M amiloride in 0.9070 NaCI solution. For home therapy, 10-3 M amiloride was prepared by the university hospital pharmacy, dissolved in physiological saline, and sealed in light-protected, coded 50-ml glass flasks. The patients or their parents withdrew 3 ml for each nebulization, and the patients inhaled it by tidal volume breathing. Solubility of the 10-3 M amiloride solution in normal saline was adequate for the 2-month test period when stored in brown, light-protected flasks. The pH of the amiloride solutions was the same as that of the vehicle. Pulmonary Function Tests/Blood Pressure Measurements Spirometry and blood pressure measurements were performed both before and after radioaerosol clearance. For the spirometry, the best of three trials was taken for the determination ofFEV t , PEFR, and IVC (inspiratory vital capacity). IVC was chosen to describe static lung volume instead of FVC because the maneuver was more reproducible and easier to perform. The blood pressure measurements served as a means to detect a possible systemic effect (i.e., a fall in BP) well known from oral administration of diuretics. Drug Delivery A Pari Standard nebulizer (Pari, Starnberg, FRO) was used for the acute amiloride therapy at the hospital, whereas Pari Boy nebulizers were used for the twice-daily therapy at home. The Pari Standard nebulizer produces an aerosol of 2.1 urn AMMD ± 1.2 (GSD) when loaded with normal saline and driven at 61pm. (Data provided by Pari.) A preliminary study showed that with the Pari Standard nebulizer charged with 3 ml of 10-3 M amiloride, an average of 0.07 mg amiloride is delivered to the airways during 10 min of inhalation by tidal volume breathing.
Mucoci/iary and Cough Clearance by Radioaerosol Technique
The deposition and clearance of an aerosol of hardened radiolabeled erythrocytes was monitored by serial scanning with a gamma camera during quiet breathing and directed coughing in order to obtain indices of mucociliary and cough clearance. The patient's own erythrocytes or blood bank (0 Rh neg) cells were labeled with 99mTc-pertechnetate, hardened with 250/0 glutaraldehyde, incubated at 60 0 e for 30 min, and thoroughly washed until the supernatant was clear. This procedure results in monodisperse tracer particles with an AMMO of 5 um (15, 16).. A Pari jet nebulizer at 6 lpm was used to produce a linear aerosol velocity of about 120cm/sthroughout the system. A copper coil (helix) following the nebulizer was designed to eliminate multiplet and fragment particles. A consecutive balloon served as a storage device and allowed further separation by means of sedimentation of multiplets. From here the patient inhaled through a combination of a stenosis and a valve system and mouthpiece, and exhaled through a filter trap. The purpose of the stenosis was to limit the inspiratory flow rate to 250 ml/s. The valve and filter system separated inhaled and exhaled aerosol and prevented radioactive contamination of the environment. The patients were instructed to breathe in slow Ive maneuvers, with a 3- to 5-s breathhold, and with a noseclip in place. The patients inhaled 30 to 40 breaths in total. This inhalation pattern was found in earlier studies to produce a more peripheral aerosol deposition and to be more reproducible than tidal volume breathing (16).The inhalation mode, total breaths and time, was noted for each patient and kept constant for all subsequent radioaerosoI inhalations in order to reproduce the initial deposition pattern for any following investigation. This procedure is essential ' because it is well known that particles in central airways clear faster by ciliary action and are more likely to clear by coughing than are particles that deposit in the periphery of the lung. If the deposition pattern after the second inhalation did not match the one from the first session, the procedure was repeated on a separate day. This turned out to be necessary for two patients, whose second clearance test was repeated for this reason. The procedure is noninvasive and welltolerated by the patient, and the radioactivity levels are modest with the inhalation of 20 to 40 fJ.Ci of radioaerosol with a half-life of 6 h. The patients received less than the equivalent of one standard thoracic. X-ray. After radioaerosol inhalation and drug or placebo inhalation the patient was seated in front of the gamma camera (Picker wide field). Thirty 2-min images were taken to measure MC as spontaneous elimination of inhaled particles. Involuntary coughing was noted and eliminated via curve correction from the MC data. The usual index of MC is the percentage of activity cleared from the lung
607
ACTION OF INHALED AMILORIDE IN CYSTIC FIBROSIS 8±3
8±5
17±7
16±8
3O±8
29±6
Results
18±7
29i8
27±9
KI8 Amilortde CJP:1.56 ± 0.78
aeutt Placebo CIP: 1.79 ± 0.53
Initial Deposition Patterns after Radloaerosollnhalatlon (% deposited)
17±8
28±9
InitialRadioaerosolDeposition Patterns
18±7
17±6
29±8
chronic PIKebo CIP: 1.68±0.73
Fig. 1. Pulmonary deposition patterns: percent deposition and C/P ratios in 23 patients in the acute trial and six patients in the chronic trial.
19±8
18±6
27±8
27±9
chronic Amilort. CJP:1.46 ± 0.84
by ciliary activity in 1 h. Regional indices of clearance (central versus peripheral) were also determined from the gamma camera images. Therefore the lung was divided in a central two thirds and a peripheral one third area of the total lung (figure 1).These defined regions from the first analysis were stored and reused for the consecutive clearance examinations. For analysis, a Gaede computer system (Gaede, Freiburg i.Brsg., FRG) was used, which allowed movement of regions of interest in the correct spot, by enlarging the images and pixel-by-pixel correction. After determining the MC, a further ten l-min images weretaken to measure CC. During the fifth minute, the patients performed a series of five standardized forced expiratory maneuvers (FET or huff), CC was determined as the difference in retained isotope activity between the first 4 min and the last 4 min. A sample of sputum for rheologic analysis was taken during this time if the patient was capable of producing one. FET or huff maneuvers were chosen because of the better reproducibility obtained than with voluntary cough.
Sputum Sampling Technique Although sputum sampling is a noninvasive technique, drawbacks include the frequent bacterial contamination and the inevitable contamination of samples with oral secretions, although we minimized this possibility by the use of dental cotton. Dental cotton was placed prior to the cough event at the openings of the salivary ducts in the mouth in order to reduce secretions and prevent possible contamination of the sputum samples (17). After each CC measurement (5 FET maneuvers), sputum was collected and weighed. Sputum rheology. The filancemeter (SEFAM, Nancy, France) measures large-amplitude elastic deformation as thread formation (18). The measurement was performed with a 20-J.1I sputum sample and a distraction ve-
locity of 10 mm/s, and generated values in millimeter units of spinnability. Sputum electrolyte content. The sputum was homogenized with 3070 acetic acid as a mucolytic agent. This suspension was centrifuged at 5,000 rpm for 15 min, and the supernatant was collected for electrolyte measurements. Sputum samples from four patients with chronic bronchitis (CBR) served as controls.
Statistical Analysis Statistical analysis was done with the aid of a Macintosh computer and the statistical program StatView 512.Data are always described as means ± SD unless otherwise indicated. Intragroup comparisons weremade by paired t testing; intergroup comparisons were made by unpaired t testing. A p value less than 0.05 was taken as significant. Study Design The study was performed in a randomized fashion for the acute and the chronic experiments. Twenty-three patients underwent an acute examination of placebo and amiloride on two separate occasions, the order ofwhicb was randomized. Six of the patients subsequently underwent a chronic inhalation trial with placebo and amiloride administered in randomized crossover fashion, with an acute examination after each 3-wk inhalation period. In each case, the acute examination included radioaerosol inhalation and clearance measurements, sputum collection, spirometry, and blood pressure measurements, as described above. The experimental analysis was carried out on a blinded basis for both the acute and chronic trials. Throughout the study, the patients were provided with coded containers for amiloride and placebo. Unfortunately, inhaled amiloride tastes slightly bitter, and most patients could readily detect the difference between the placebo and the active drug.
The radioaerosol deposition patterns obtained in our patients are illustrated in figure 1. The data are expressed as percent regional deposition as well as the central/peripheral ratio (C/P). The variations in deposition pattern between different study days were not statistically significant. Thus, differences in initial deposition pattern can be ruled out as a possible explanation for variations in bronchial clearance.
Mucociliary and Cough Clearance Acuteeffects: mucociliary clearance. The acute effect of amiloride on MC is illustrated in figure 2, as wellas the individual changes in each patient at 30 and at 60 min. The overall MC in our 23 patients, expressed as the percent of initial deposition cleared by ciliary action in 1 h, was 29.1 ± 8.2070 for placebo and 35.8 ± 10.4070 for amiloride. This result confirms the conclusion of our earlier study (13)that amiloride is able to improve MC in patients with CF, but the significance level is now much higher than previously reported (p = 0.0004 versus 0.002 at 30 min and p = 0.0007 versus 0.005 at 60 min of MC). The placebo MC values in the patients with CF weregreater than those reported for smokers with simple chronic bronchitis (19 ± 8070) but lower than observed normal values at 20 yr of
• o
•
0 90
Placebo
80
80
• Amlloride
70
70
60
60
50
50
40
40
100
C
95
E
90
0
I II:
30
85
!Q
80
c
= 5
75
4D
70
.E 65 ~ 60 .2
J
N: 23 CF patients
55
50
~ 0
i
i
i
i
i
i
10
20
30
40
50
60
Time in minutes
Fig. 2. Mucociliary clearance and individual clearance rates at 30 and 60 min for the 23 patients in the acute trial. *p = 0.0004; **p = 0.0007.
608
App, KING, HELFESRIEDER, KOHLER, AND MATTHYS
18
f:§:1Placebo
• Amiloride
16
parameters such as body height, weight, clinical status, or treatment. Chroniceffects:coughclearance. The baseline CC and the increase in CC caused by a simple acute challenge with amiloride were also comparable between the chronic subgroup and the overall group, although the acute amiloride effect was nonsignificant. Also, the chronic placebo inhalation was not significantly different from the acute placebo inhalation (8.2 ± 3.4% chronic, 8.8 ± 3.4% acute). Nevertheless, the chronic amiloride inhalation improved the standardized cough effectiveness markedly (p = 0.004) to a value of 16.8 ± 4.2% versus 11.3 ± 7.0% for acute amiloride.
Age: 15.9 ! 6.2 yn.
40
3.
N: 23 CF patients
30
25
iii
2
ERNST M. APP, MALCOLM KING, RENATE HELFESRIEDER, DIETER KOHLER, and HEINRICH MATTHYS
Introduction Cystic fibrosis (CF), first described just 50 yr ago, is today the most common lethal inherited disease among Caucasians (1).This autosomal recessive disorder has a carrier rate ranging from approximately 1:20 in most of Europe and among white Americans to 1:60 in U.S. blacks and 1:150in the Japanese of Hawaii. The basic defect relates to regulation of ion transport by the secretory epithelia. A reduced chloride conductance occurs in both airway and sweat ductal epithelia in CF (2). In addition, in airway epithelium, but not in sweat gland, the sodium absorption is abnormally elevated (2, 3). Thus, if augmented volume absorption and/or reduced water secretion playa role in the pathogenesis of CF airway disease, then modalities to either induce CI- secretion or block the excess Na' absorption may be beneficial. Because no established drug activates the defective CI- channels in CF airway epithelia, the alternative concept, i.e., blocking Na" absorption, appears to be the most rational approach to develop a useful therapy in CF. Investigations have shown that bronchial clearance is impaired in patients with CF (4). This may be due to rheologic abnormalities in the CF mucus, which could arise from dysfunction of transepithelial electrolyte movement (5). In CF, the respiratory secretions exhibit decreased sodium and chloride contents (6), only partially compensated by increased potassium and other electrolytes, which appear to reflect the altered ion transport mechanisms. Whether this results in a decreased water content and hence increased viscosity or rigidity is not yet certain, although the water content of CF sputum (~ 89010) has been reported to be lower than in other lung diseases (~ 95010) (7) or than that of normal canine tracheal mucus (~ 93070) (8, 9). Thus, alterations in electrolyte content might corre-
SUMMARY Cystic fibrosis (CF) Is the most common Inherited fatal disorder among caucasians. Bronchial mucus In CF contains more potassium and less sodium, which may be due to Increased sodium absorption, resulting In a reduced airway water content. We studied 23 Patients with CF after Inhalation of normal saline or amllorlde (10-3 M), a sodium transport bloc.r. Mucocillary clearance (MC) and cough clearance (CC) were determined with a gamma camera that traced the movement of """fc-Iabeled, hardened erythrocytes over a 1-h Period after the patients Inhaled these particles as an aerosol. Before and after each Investigation pulmonary function tests (PFT) and blood pressure (BP) were measured. Sputum thread formation was measured by means of a fllancemeter. Six of the patients also completed a 3-wk trial of amllorlde Inhalation therapy. MC Increased significantly (p < 0.001) after acute amllorlde Inhalation (bronchial deposition, O.fIT mg amllorlde) compared with that In the saline control. CC also Increased, but not as much as MC.After 3 wk of amlloride Inhalation (2 times a day) clearance values (both MC and CC) were markedly enhanced (p < 0.01); after a similar Period of saline Inhalation, clearance values were not different from baseline. Sputum fllance values also decreased significantly after amllorlde Inhalation. There were no adverse effects of the amllorlde Inhalation compared with saline. We conclude that amllorlde Inhal. tlon administered as a single dose or as long-term therapy Is able to Incre_ MC and CC In CF airways and that the effect of 10-3 M amllorlde Inhalation on MC lasts at least 40 min. These results Indicate that further clinical studies of long-term amllorlde Inhalation In Patients with CF should be undertaken. This Is the first line of evidence that an Inhaled diuretic drug such as amllorlde Is able to enhance mucus clearance favorably In patients with CF. The mechanism Is most lliely due to the alteration of water content within the airways. AM REV RESPIR DIS 1990; 141:605-612
late with alterations in sputum rheology, which in tum might relate to impairments in ciliary and/or cough clearability of mucus. These suggestions led us to study amiloride (pyrazinecarbonylguanidine) as a novel therapeutic approach to this disease. Boucher and coworkers (3) and Knowles and colleagues (10) had previously shown that amiloride applied directly to the luminal side of the respiratory epithelium is able to "normalize" the increased transepithelial potential difference in patients with CF, and they proposed its use as a therapy in CF lung disease. Amiloride as a diuretic drug acts on the nephron in the kidney by blocking sodium reabsorption while preventing potassium excretion. For this to occur the amiloride must first be filtered in the glomerulus and act on the luminal side of the epithelium (11). A similar mechanism could result from the inhalation of amiloride with a reduction of the transepithelial potential difference and an in-
crease in the amount of sodium in pulmonary secretions. Curtis and Misch (12) showed a decreased mucociliary transport on the frog palate in vitro after amiloride, and they suggested this may be due to increased mucus discharge and altered mucus rheology. However, in a preliminary report (13), we were able to show an increase
(Received in original form April 10, 1989 and in revised form August 16, 1989) 1 From the Pulmonary Defense Group, University of Alberta, Edmonton, Alberta, Canada; the Division of Pulmonary Disease, Department of Internal Medicine, University of Freiburg, Freiburg i.Brsg., and the Fachkrankenhaus Kloster Grafschaft, Schmallenbrg, Federal Republic of Germany. 2 Supported by the Ministry for Science and Art of the Federal State of Baden-Wiirttemberg and by the Canadian Cystic Fibrosis Foundation. 3 Correspondence and requests for reprints should be addressed to Ernst M. App, M.D., Pulmonary Defense Group, 519 Newton Research Bldg, University of Alberta, Edmonton, Alberta TOO 2C2, Canada.
605
AP~ KING, HELFESRIEDER, KOHLER, AND MATTHYS
606
in mucus clearance as an acute effect of amiloride in CR Recently, Benedetto and coworkers (14)presented their results on cilia beat frequency in vitro and found a small but statistically significant increase in cilia beat frequency (CBF) with amiloride for approximately 1 h, which may account for the increased mucociliary clearance (MC) in CF that we previously demonstrated. We therefore wanted to examine the mechanisms for the acute improvement of MC, as well as cough clearance (CC), in CF after inhalation of amiloride. Furthermore, wewanted to explore the benefits of longer term or chronic amiloride inhalation on bronchial mucus clearance in such patients. Methods Patients I\venty-three patients (16 male and seven female) participated in the acute trial of amiloride therapy, including 14 patients described in the previous report (13).They were recruited from the outpatient clinic in Freiburg and from three other cystic fibrosis centers in the southwest of Germany. Their mean age (± SD) was 15.9 (± 6.2) with a range of 7 to 27 yr. Their height varied between 119and 181 em (mean, 153.9 ± 20.4 em) and their weight between 18and 65 kg (mean, 39.8 ± 13.4kg). They represented the whole spectrum from mild to severe symptomatic, from primarily gastrointestinal symptoms to mainly pulmonary, although no patients with acute exacerbations were studied. Of these 23 patients, six (three male and three female) subsequently underwent a 3-wk twice-daily inhalation trial with 10-3 M amiloride and the same period for placebo (0.9070 NaCI). The mean age (± SO) in this group was 17.1 (± 4.9) with a range of 12 to 23 yr. Their body height was between 139 and 179 em with an average of 161.3 (± 14.9), and a weight range between 23 and 54 kg with a mean of 41.3 (± 11.5). On the basis of their experience with the acute trial, the patients decided for themselves if they wanted to undertake this long-term therapy trial and were then entered at random in the study design. A possible bias may be that only "responder" patients decided to participate, but this did not turn out to be the case (see below). For each investigation the patient and/or the accompanying parent was informed about the procedure and the possible risks involved, and the consent was given in writing. The study was approved by the institutional ethics committee. Inclusion criteria. Confirmed diagnosis of CF. The diagnosis of CF in our patients was confirmed by typical history, clinical, radiographic, and sweat test findings. Exclusion criteria. (1) Pregnancy. (2) Patients with an acute pulmonary infection. (3) Patients younger than 7 yr of age, because
cooperation is required for aerosol inhalation and follow-up studies. Therapy changes. For the measurement of the acute effect of amiloride, we ensured that the patients did not alter their ongoing therapy between the two occasions of mucus clearance examination, which occurred on two separate days (at the same time of day) not longer than 1 wk apart. For the study of the chronic effect of amiloride, the patients were also maintained on their usual therapy (chest physiotherapy, inhaled sympathomimetics) except for mucolytic agents, which were withdrawn and replaced with amiloride or saline. None of the patients was receiving oral antibiotics during the chronic trial.
Drug Background and Concentration A dose-dependent effect of amiloride on airway bioelectric properties has been described over the concentration range of 10-7 to 10-3 M (12). Higher concentrations of drug, on the order of 10-2 M, are not soluble in physiological saline without reducing the overall electrolyte concentration. For the acute aerosol measurements, we chose to nebulize freshly prepared 10-3 M amiloride in 0.9070 NaCI solution. For home therapy, 10-3 M amiloride was prepared by the university hospital pharmacy, dissolved in physiological saline, and sealed in light-protected, coded 50-ml glass flasks. The patients or their parents withdrew 3 ml for each nebulization, and the patients inhaled it by tidal volume breathing. Solubility of the 10-3 M amiloride solution in normal saline was adequate for the 2-month test period when stored in brown, light-protected flasks. The pH of the amiloride solutions was the same as that of the vehicle. Pulmonary Function Tests/Blood Pressure Measurements Spirometry and blood pressure measurements were performed both before and after radioaerosol clearance. For the spirometry, the best of three trials was taken for the determination ofFEV t , PEFR, and IVC (inspiratory vital capacity). IVC was chosen to describe static lung volume instead of FVC because the maneuver was more reproducible and easier to perform. The blood pressure measurements served as a means to detect a possible systemic effect (i.e., a fall in BP) well known from oral administration of diuretics. Drug Delivery A Pari Standard nebulizer (Pari, Starnberg, FRO) was used for the acute amiloride therapy at the hospital, whereas Pari Boy nebulizers were used for the twice-daily therapy at home. The Pari Standard nebulizer produces an aerosol of 2.1 urn AMMD ± 1.2 (GSD) when loaded with normal saline and driven at 61pm. (Data provided by Pari.) A preliminary study showed that with the Pari Standard nebulizer charged with 3 ml of 10-3 M amiloride, an average of 0.07 mg amiloride is delivered to the airways during 10 min of inhalation by tidal volume breathing.
Mucoci/iary and Cough Clearance by Radioaerosol Technique
The deposition and clearance of an aerosol of hardened radiolabeled erythrocytes was monitored by serial scanning with a gamma camera during quiet breathing and directed coughing in order to obtain indices of mucociliary and cough clearance. The patient's own erythrocytes or blood bank (0 Rh neg) cells were labeled with 99mTc-pertechnetate, hardened with 250/0 glutaraldehyde, incubated at 60 0 e for 30 min, and thoroughly washed until the supernatant was clear. This procedure results in monodisperse tracer particles with an AMMO of 5 um (15, 16).. A Pari jet nebulizer at 6 lpm was used to produce a linear aerosol velocity of about 120cm/sthroughout the system. A copper coil (helix) following the nebulizer was designed to eliminate multiplet and fragment particles. A consecutive balloon served as a storage device and allowed further separation by means of sedimentation of multiplets. From here the patient inhaled through a combination of a stenosis and a valve system and mouthpiece, and exhaled through a filter trap. The purpose of the stenosis was to limit the inspiratory flow rate to 250 ml/s. The valve and filter system separated inhaled and exhaled aerosol and prevented radioactive contamination of the environment. The patients were instructed to breathe in slow Ive maneuvers, with a 3- to 5-s breathhold, and with a noseclip in place. The patients inhaled 30 to 40 breaths in total. This inhalation pattern was found in earlier studies to produce a more peripheral aerosol deposition and to be more reproducible than tidal volume breathing (16).The inhalation mode, total breaths and time, was noted for each patient and kept constant for all subsequent radioaerosoI inhalations in order to reproduce the initial deposition pattern for any following investigation. This procedure is essential ' because it is well known that particles in central airways clear faster by ciliary action and are more likely to clear by coughing than are particles that deposit in the periphery of the lung. If the deposition pattern after the second inhalation did not match the one from the first session, the procedure was repeated on a separate day. This turned out to be necessary for two patients, whose second clearance test was repeated for this reason. The procedure is noninvasive and welltolerated by the patient, and the radioactivity levels are modest with the inhalation of 20 to 40 fJ.Ci of radioaerosol with a half-life of 6 h. The patients received less than the equivalent of one standard thoracic. X-ray. After radioaerosol inhalation and drug or placebo inhalation the patient was seated in front of the gamma camera (Picker wide field). Thirty 2-min images were taken to measure MC as spontaneous elimination of inhaled particles. Involuntary coughing was noted and eliminated via curve correction from the MC data. The usual index of MC is the percentage of activity cleared from the lung
607
ACTION OF INHALED AMILORIDE IN CYSTIC FIBROSIS 8±3
8±5
17±7
16±8
3O±8
29±6
Results
18±7
29i8
27±9
KI8 Amilortde CJP:1.56 ± 0.78
aeutt Placebo CIP: 1.79 ± 0.53
Initial Deposition Patterns after Radloaerosollnhalatlon (% deposited)
17±8
28±9
InitialRadioaerosolDeposition Patterns
18±7
17±6
29±8
chronic PIKebo CIP: 1.68±0.73
Fig. 1. Pulmonary deposition patterns: percent deposition and C/P ratios in 23 patients in the acute trial and six patients in the chronic trial.
19±8
18±6
27±8
27±9
chronic Amilort. CJP:1.46 ± 0.84
by ciliary activity in 1 h. Regional indices of clearance (central versus peripheral) were also determined from the gamma camera images. Therefore the lung was divided in a central two thirds and a peripheral one third area of the total lung (figure 1).These defined regions from the first analysis were stored and reused for the consecutive clearance examinations. For analysis, a Gaede computer system (Gaede, Freiburg i.Brsg., FRG) was used, which allowed movement of regions of interest in the correct spot, by enlarging the images and pixel-by-pixel correction. After determining the MC, a further ten l-min images weretaken to measure CC. During the fifth minute, the patients performed a series of five standardized forced expiratory maneuvers (FET or huff), CC was determined as the difference in retained isotope activity between the first 4 min and the last 4 min. A sample of sputum for rheologic analysis was taken during this time if the patient was capable of producing one. FET or huff maneuvers were chosen because of the better reproducibility obtained than with voluntary cough.
Sputum Sampling Technique Although sputum sampling is a noninvasive technique, drawbacks include the frequent bacterial contamination and the inevitable contamination of samples with oral secretions, although we minimized this possibility by the use of dental cotton. Dental cotton was placed prior to the cough event at the openings of the salivary ducts in the mouth in order to reduce secretions and prevent possible contamination of the sputum samples (17). After each CC measurement (5 FET maneuvers), sputum was collected and weighed. Sputum rheology. The filancemeter (SEFAM, Nancy, France) measures large-amplitude elastic deformation as thread formation (18). The measurement was performed with a 20-J.1I sputum sample and a distraction ve-
locity of 10 mm/s, and generated values in millimeter units of spinnability. Sputum electrolyte content. The sputum was homogenized with 3070 acetic acid as a mucolytic agent. This suspension was centrifuged at 5,000 rpm for 15 min, and the supernatant was collected for electrolyte measurements. Sputum samples from four patients with chronic bronchitis (CBR) served as controls.
Statistical Analysis Statistical analysis was done with the aid of a Macintosh computer and the statistical program StatView 512.Data are always described as means ± SD unless otherwise indicated. Intragroup comparisons weremade by paired t testing; intergroup comparisons were made by unpaired t testing. A p value less than 0.05 was taken as significant. Study Design The study was performed in a randomized fashion for the acute and the chronic experiments. Twenty-three patients underwent an acute examination of placebo and amiloride on two separate occasions, the order ofwhicb was randomized. Six of the patients subsequently underwent a chronic inhalation trial with placebo and amiloride administered in randomized crossover fashion, with an acute examination after each 3-wk inhalation period. In each case, the acute examination included radioaerosol inhalation and clearance measurements, sputum collection, spirometry, and blood pressure measurements, as described above. The experimental analysis was carried out on a blinded basis for both the acute and chronic trials. Throughout the study, the patients were provided with coded containers for amiloride and placebo. Unfortunately, inhaled amiloride tastes slightly bitter, and most patients could readily detect the difference between the placebo and the active drug.
The radioaerosol deposition patterns obtained in our patients are illustrated in figure 1. The data are expressed as percent regional deposition as well as the central/peripheral ratio (C/P). The variations in deposition pattern between different study days were not statistically significant. Thus, differences in initial deposition pattern can be ruled out as a possible explanation for variations in bronchial clearance.
Mucociliary and Cough Clearance Acuteeffects: mucociliary clearance. The acute effect of amiloride on MC is illustrated in figure 2, as wellas the individual changes in each patient at 30 and at 60 min. The overall MC in our 23 patients, expressed as the percent of initial deposition cleared by ciliary action in 1 h, was 29.1 ± 8.2070 for placebo and 35.8 ± 10.4070 for amiloride. This result confirms the conclusion of our earlier study (13)that amiloride is able to improve MC in patients with CF, but the significance level is now much higher than previously reported (p = 0.0004 versus 0.002 at 30 min and p = 0.0007 versus 0.005 at 60 min of MC). The placebo MC values in the patients with CF weregreater than those reported for smokers with simple chronic bronchitis (19 ± 8070) but lower than observed normal values at 20 yr of
• o
•
0 90
Placebo
80
80
• Amlloride
70
70
60
60
50
50
40
40
100
C
95
E
90
0
I II:
30
85
!Q
80
c
= 5
75
4D
70
.E 65 ~ 60 .2
J
N: 23 CF patients
55
50
~ 0
i
i
i
i
i
i
10
20
30
40
50
60
Time in minutes
Fig. 2. Mucociliary clearance and individual clearance rates at 30 and 60 min for the 23 patients in the acute trial. *p = 0.0004; **p = 0.0007.
608
App, KING, HELFESRIEDER, KOHLER, AND MATTHYS
18
f:§:1Placebo
• Amiloride
16
parameters such as body height, weight, clinical status, or treatment. Chroniceffects:coughclearance. The baseline CC and the increase in CC caused by a simple acute challenge with amiloride were also comparable between the chronic subgroup and the overall group, although the acute amiloride effect was nonsignificant. Also, the chronic placebo inhalation was not significantly different from the acute placebo inhalation (8.2 ± 3.4% chronic, 8.8 ± 3.4% acute). Nevertheless, the chronic amiloride inhalation improved the standardized cough effectiveness markedly (p = 0.004) to a value of 16.8 ± 4.2% versus 11.3 ± 7.0% for acute amiloride.
Age: 15.9 ! 6.2 yn.
40
3.
N: 23 CF patients
30
25
iii
2
![[Pilot study of amiloride inhalation in children with cystic fibrosis].](/assets/img/hold.png)
