INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Pulmonary disease in cystic fibrosis (CF) is characterized by production of an abnormal mucus, airway obstruction, and bacterial colonization. This leads to progressive deterioration of pulmonary function (1). Besides genetically regulated effects in transepithelial sodium and chloride transport, high concentrations of DNA characterize CF mucous abnormalities (2, 3). Recombinant human deoxyribonuclease (rhDNase I), a drug able to digest DNA reduces the viscosity of CF sputum (4, 5), improves pulmonary function, and reduces the frequency of exacerbations requiring antibiotic therapy (6). Most patients are treated with a single nebulized dose per day. However, there are in vitro data demonstrating effects of local concentration of rhDNase I on mucous rheology (4, 5), suggesting that some patients may benefit from increases in daily therapy.

It is generally known that the lung dose of an aerosolized drug is not directly related to the nebulized dose but there are few data available in younger patients. Nebulizer efficiency, particle distribution, breathing pattern and airway pathophysiology can affect deposition and clinical response. Because of differences in physiochemical properties between drugs and pathophysiological changes in patients with different diseases, it is difficult to predict the lung dose and response for a given aerosolized drug (11, 12, 14). For rhDNase I, the three recommended nebulizers (Acorn II; Marquest, Englewood, CA. Pari LC jet; PARI Respiratory Equipment Inc., Richmond VA. T UpDraft; Hudson RCI, Temecula, CA) have been matched for aerosol production and particle distribution (13), but the particles are relatively large and upper airway deposition may cause laryngeal irritation (8).

To better understand in vivo factors influencing lung dose of aerosolized rhDNase I, we measured deposition in 15 clinically stable patients with CF who were on chronic daily therapy. Total and regional deposition were correlated with breathing pattern, pulmonary function, demographic factors and disease severity. In addition, the effects of each patient's measured lung dose on pulmonary function was estimated by stopping the drug and observing changes in spirometry over a 2-wk follow-up period. The 2-wk delay was chosen as a compromise allowing an effective washout of rhDNase I to detect possible clinical effects with limited risk of acute exacerbation. Although the mechanism of action of rhDNase I in vivo remains uncertain, it is now widely accepted from previous placebo controlled studies and from clinical experience that patients feel better while on the drug and pulmonary function is improved. Therefore, this study was not designed to be blinded nor to determine efficacy but to determine possible dose- response relationships that influence efficacy.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Principles

Because we aimed to assess factors influencing rhDNase I aerosol deposition under the identical conditions of home therapy, all experiments were performed without attempts to control patient related variables, allowing each patient to treat him or herself as if they were at home. Patients entered the study while on the drug in a clinical steady state. They inhaled a full dose of rhDNase I using their own nebulizer/compressor combination, which took between 10 and 15 min. The breathing patterns during rhDNase I were carefully measured. Immediately after rhDNase I aerosol inhalation had been completed, a surrogate radiolabeled aerosol was administered using the same nebulizer for 1 to 2 min. This timing was short enough to prevent mucociliary clearance effects from disturbing the distribution of the deposited particles (the patients did not cough during this period). The breathing pattern during the surrogate aerosol inhalation was carefully monitored and verified to be similar to the rhDNase I aerosol inhalation. Immediately following the deposition study, therapy with rhDNase I was discontinued for 2 wk. Spirometry was measured on the day of the deposition study and 2 wk later.

Patients

Fifteen patients (five males and 10 females) diagnosed with CF were enrolled by signing a consent form approved by the Hospital and University Human Studies Committees. Patients had to fulfill all of the following inclusion criteria: no hospitalization, intravenous antibiotics or change in baseline therapy for 2 wk prior to entry, no hemoptysis greater than 30 ml during the previous month, no hemoptysis requiring embolization or transfusion in the past 6 mo, no more than a 10% change in FEV₁ from most recent formal testing. Ages ranged between 6 and 31 yr with a mean of 16.9 ± 1.8 (mean ± SE). Six patients were homozygous for the Delta F 508 (DF 508) mutation, and five were heterozygous. Shwachman score, consistently measured by one of us (JD-G) ranged between 65 and 97 (79.6 ± 2.40, mean ± SE). Twelve patients presented with chronic bacterial colonization with either Pseudomonas aeruginosa or Staphylococcus aureus. All patients had received chronic therapy with rhDNase I for months since the initial FDA approval, using the recommended nebulizers and compressors. Eleven used the Acorn II (Marquest) jet nebulizer with PulmoAide compressor (DeVilbiss, Somerset, PA); three used the Pari LC jet nebulizer with the Pari Proneb compressor (PARI Respiratory Equipment Inc.); and one patient used the T UpDraft nebulizer (Hudson RCI) with PulmoAide compressor. Patients using the Acorn II or T UpDraft changed the nebulizer each day. With Pari LC jet, patients used the same nebulizer each day, cleaning it after each treatment as recommended. Fourteen patients were treated with rhDNase I at a daily dose of 2.5 mg, one patient 2.5 mg twice a day. Prior to treatment with rhDNase I, all patients performed their routine pulmonary toilet, including any nebulized bronchodilator therapy and pulmonary drainage.

Aerosol Bench Studies

Using a bench protocol previously described from our laboratory, both standard rhDNase I and surrogate aerosols were generated with all three nebulizers. To characterize rhDNase I aerosols for each nebulizer, we added 0.1 ml of technetium pertechnetate (^99mTc) in saline to the standard 2.5 ml solution containing 2.5 mg of the drug; this solution generated our aerosol standards. For human radiolabel studies, the surrogate aerosol consisted of 2.5 ml of normal saline mixed with ^99mTc bound to human serum albumin (^99mTc-HSA; 21 mg Medi-Physics, Inc., Arlington Heights, IL) to avoid passage of the ^99mTc into the blood.

Nebulizer function and particle distributions for all aerosols were quantitated using a piston ventilator that mimics the clinical delivery of nebulized medication to a spontaneously breathing patient and a cascade impactor (GS 1; California Measurements, Sierra Madre, CA [14-17]). The conditions of nebulization were fixed at a tidal volume of 500 ml, breathing frequency of 15/min and duty cycle of 0.5. Aerosolized particles were captured on filters and quantified by measuring radioactivity in a well counter. To demonstrate the accuracy of the surrogate aerosol in duplicating the behavior of the rhDNase I aerosols, for each nebulizer, we plotted the relationship between radioactivity for the captured aerosolized rhDNase I particles versus the ^99mTc-HSA saline particles from the cascade impactor as a line of identity (18). Further, the aerodynamic behavior (mass median aerodynamic diameter [MMAD] and geometric standard deviation [g]) of the sampled aerosols was quantitated on probability paper.

Aerosol Deposition Imaging

On the study day, the patient performed routine pulmonary toilet/ bronchodilator therapy and spirometry was measured. The patient was seated in front of a posteriorly positioned gamma camera (Picker Dyna camera; Northford, CT) initially set for zenon (¹³³Xe). Breathing tidally at functional residual capacity (FRC) the patient inhaled 5-10 mCi of ¹³³Xe until the count rate became stable ± 10% over 15 s. One min gamma camera images (70,000-100,000 counts) were acquired and stored in a computer for analysis (Advanced Medical Computer, Sunnyvale, CA). After ¹³³Xe equilibrium, the camera was switched to ^99mTc. Then, using his/her own nebulizer and compressor, the patient inhaled a full dose of aerosolized rhDNase I. During inhalation, flow was measured via a pneumotachograph placed on the expiratory line of the nebulizer, integrated over time to yield volume and recorded on a strip chart recorder. These tracings provided flow, tidal volume, frequency, and duty cycle for the complete treatment period which would take from 10-15 min.

Immediately following rhDNase I deposition, the nebulizer was washed, dried and refilled with 2.5 ml normal saline labeled with 5 mCi of ^99mTc-HSA. The patient inhaled the ^99mTc HSA aerosol until the activity deposited in the thorax, continuously monitored over 15 s periods of time reached approximately 10,000 counts. The total time of ^99mTc-HSA inhalation, which took between 1 and 2 min, was carefully monitored. The breathing pattern was continuously recorded as described above. Immediately following inhalation, and after a drink of water to wash activity out the pharynx into the stomach, a 60 s deposition image was obtained.

Quantification of Aerosol Deposition

We measured total aerosol deposition using a mass balance technique. With this method, the amount of radioactivity inhaled by the patient and the amount exhaled are measured using filters (12, 15, 16). Briefly, the technique involves two measurements, the in vivo treatment run followed by the in vitro calibration run.

Treatment Run

During the patient inhalation of the ^99mTc-HSA saline aerosol a filter was placed on the exhalation line of the nebulizer which captured all the particles exhaled by the patient in addition to the particles that passed from the nebulizer to the filter which were not inhaled (i.e., the leakage). The leakage occurs primarily during expiration because the nebulizer runs continuously.

Calibration Run

After deposition was completed, the nebulizer containing the remaining ^99mTc-HSA saline solution was installed in an in vitro testing circuit. During nebulization, the mouthpiece was attached to a piston pump (Harvard Apparatus, South Natick, MA), which duplicated each patient's breathing pattern (average tidal volume, frequency, duty cycle and treatment time from the treatment run [13]). Two filters were used during the calibration run, the inhalation filter and the leakage filter. The inhalation filter was placed between the mouthpiece and the pump to capture all particles that would have been inhaled by the patient. The leakage filter placed on the exhalation line measured the leakage because no particles were exhaled from the inhalation filter. Thus, the inhalation filter provides the amount of aerosol inhaled by a given patient and the difference between the leakage filter and the exhalation filter used during the treatment run determines the amount exhaled by the patient. The fraction of the radiolabeled aerosol deposited in the patient is given by Equation 1.

(1)Deposition Fraction (DF) = [Radioactivity Inhaled  Radioactivity Exhaled]/Radioactivity Inhaled

Knowledge of the DF is useful because it represents the fraction of aerosol inhaled that deposited in the patient. If the amount of rhDNase I inhaled is known (inhaled mass), then rhDNase I deposition is given by Equation 2.

(2)rhDNase I deposition = [Inhaled Mass of rhDNase I] × DF

The inhaled mass was measured in a final bench experiment. The patient's own nebulizer was washed and refilled with saline labeled with ^99mTc-HSA. Using the piston ventilator the average breathing pattern measured during rhDNase I inhalation was duplicated. With the inhalation filter in place, the inhaled mass was determined by running the nebulizer for the same time the patient was actually treated with rhDNase I. When expressed as a percentage of the initial amount of radioactivity placed in the nebulizer, the inhaled mass in mg is easily calculated by multiplying the percentage radioactivity on the filter by the number of mg of drug in the nebulizer (12, 15, 16).

Analysis

Regional deposition. The mass balance technique described above measures dose deposited in the patient. To determine the regional distribution of the dose (e.g., right versus left lung, upper airways, etc.) gamma camera imaging of deposited radioactivity was utilized.

We expressed the distribution of deposited particles in terms of regions of interest generated by the computer. Regions were drawn over the ¹³³Xe equilibrium scan; a region over the entirety of both lungs called the whole lung zone, and another region centered over the large central airways comprising 33% of the area of both lungs, which we called the central zone. The area remaining after the central zone was deducted from the whole lung zone was called the peripheral zone (19). The ¹³³Xe regions of interest were then superimposed over the ^99mTc deposition image. The ratio between the central (C) and peripheral (P) lung counts (sC/P) was calculated in a manner which normalized for differences in relative lung thickness by dividing the C/P ^99mTc counts by the C/P ¹³³Xe counts. This ratio defined the specific C/P ratio (sC/P) (19). Using the resulting sC/P values, a ratio of 1.0 reflects deposition proportional to regional volume. Because the central region outlines both central airways and the lung parenchyma surrounding them, an sC/P ratio of unity reflects predominantly alveolar deposition. Increasing sC/P ratios greater than unity indicate increasing deposition in the proximal airways. Radioactivity deposited in the pharynx and upper airways and/or swallowed was estimated by counting over the stomach after a drink of water.

Dose-response between lung deposition of rhDNase I and spirometric changes was assessed by plotting the change in FEV₁ as a percent of the baseline (on therapy) versus the mass of drug deposited in the lungs. The latter term was obtained by multiplying the total deposition of drug in mg by the percentage of lung counts measured by the regional analysis of the deposition image.

Statistical Tests

Statistical group comparisons for spirometry, breathing pattern and deposition data were made using paired analysis. Correlations were assessed by linear and non-linear regression.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Bench Studies

Regression equations describing the aerodynamic behavior of rhDNase I and saline-HSA aerosols for the three nebulizers are listed in Table 1. For each nebulizer there was a close correlation between the rhDNase I aerosols and the surrogate saline-HSA.

Particle distributions for each aerosol are shown in Figure 1, a plot of log aerogynamic diameter versus probability. The data for all the aerosols listed in Table 1 are obviously superimposed and aerodynamically indistinguishable. Assuming a log-normal distribution, a best fit straight line reveals a MMAD of 3.0 µm with a g of 2.3. 

View larger version (14K): [in this window] [in a new window]   Figure 1.   Particle distribution of rhDNase I (open symbols) and human serum albumin (HSA) in normal saline (closed symbols) aerosols sampled by cascade impaction. Data are plotted as the log aerodynamic diameter versus probability, points to the right of the line represent the absolute filter downstream to the cascade impactor (MMAD = 3.0 µm, g = 2.3).

Patient Data

All 15 patients completed the study. Subjectively, all patients described worsening of dyspnea and increased difficulty in producing sputum during the two week washout period. Objective data are listed in Table 2. In 12 patients, there was a decrease in FEV₁ when compared to baseline values measured before stopping the drug (mean FEV₁ % predicted ± SE on rhDNase I = 86.9 ± 5.6%, and 77.8 ± 5.7% off, p = 0.003). There was an inverse correlation between age and Shwachman score (r = 0.64, p < 0.05). Descriptive data for the patient breathing patterns for DNase inhalation are also listed. Paired comparisons with the HSA surrogate aerosol inhalation were not significantly different.

Total Deposition

Total aerosol deposition measured by mass balance varied significantly from patient to patient (Table 3) with a range of 0.160 to 0.850 mg rhDNase I deposited. Linear regression analysis of the factors determining deposition (inhaled mass and DF from Equation 2) indicated that more of the variation was related to the deposition fraction (DF) (mean ± SE of 0.818 ± 0.057, r = 0.861, p < 0.0001) than the inhaled mass (27.5 ± 1.14%, r = 0.573, p = 0.026). DF did not correlate significantly with any breathing parameter listed in Table 2, but did correlate with the ratio of FEV₁/FVC % (r = 0.58, p < 0.05) such that those patients with more obstruction deposited a greater fraction of inhaled drug.

Regional Deposition

Total patient deposition was partitioned into whole lung and upper airway deposition (Table 3). In some patients, as much as 48% of the deposited aerosol was found in the upper airways (range 0.0 to 0.30 mg). An example is shown in Figure 2 (left panel ). Patient 4, 9 yr old, exhibited stomach activity at 48% of the total after the pharyngeal wash. During rapid inhalation, upper airway deposition is known to correlate positively with inspiratory flow (19), however, we found a negative correlation in our patients (r = 0.692, p = 0.013) who were breathing tidally. When breathing patterns were further analyzed, a negative correlation with tidal volume was found (Figure 3; r = 0.696, p < 0.006). These observations indicated that upper airway deposition may be associated with factors related to growth, and, as shown in Figure 4, upper airway deposition was negatively correlated with age (r = 0.743, p < 0.005).

View larger version (114K): [in this window] [in a new window]   Figure 2.   Gamma camera images following rhDNase I deposition. (Left panel ) 9-yr-old boy. Upper airway deposition represented by stomach activity 48% of total, uniform deposition observed over the lungs. (Right panel ) 31-yr-old woman. Upper airway deposition is much reduced (4.3%), lung deposition is nonuniform with patchy airway activity.

View larger version (12K): [in this window] [in a new window]   Figure 3.   Upper airway deposition (% of total) versus tidal volume (y = 0.94707  47.460*LOG(x); r = 0.696, p < 0.006).

View larger version (10K): [in this window] [in a new window]   Figure 4.   Upper airway deposition (% of total) versus age (y = 75.434  50.212*LOG(x); r = 0.743, p < 0.005).

For the lungs, deposition ranged between 0.16 mg and 0.78 mg (0.47 ± 0.04 mg). Figure 2 (right panel) depicts a typical scan from an older patient (#1, 31 yr of age). There was much less stomach activity indicating reduced upper airway deposition (4.3% of total). When compared to the younger patient, the pattern of deposition in the lungs was more irregular with an increased tendency for patchy deposition in central airways. Overall, lung deposition was significantly related to FEV₁ as shown in Figure 5, with deposition increasing as lung function decreased (r = 0.611, p = 0.0152). Analysis of regional deposition within the lungs by sC/P ratios are shown for each lung in Table 3. Reflecting the range of disease, there was a wide range in lung sC/P with some patients approximating 1.0 indicating peripheral deposition and others at greater than 2.0 consistent with significant airways disease and concomitant central deposition. But, while the aerosol tended to deposit more centrally in the airways in the older patients (r = 0.428 with age) and those with more disease (r = 0.488 with FEV₁%) those correlations were not significant. Figure 6 demonstrates the relationship between drug effect, as assessed by changes in FEV₁ after cessation of therapy, and measured lung dose. There was no significant correlation (r = 0.066, p = 0.816). Further, when changes in FEV₁ were related to the dose in each lung adjusted for the sC/P ratio of that lung there was no correlation between dose and response.

View larger version (12K): [in this window] [in a new window]   Figure 5.   Lung deposition versus FEV1 (% predicted; r = 0.611, p = 0.0152).

View larger version (12K): [in this window] [in a new window]   Figure 6.   Decrements in FEV1 after cessation of therapy as a percentage of baseline versus lung dose of rhDNase I (lung dose for Patient 1 doubled from the value listed in Table 3 because of bid regimen). No correlation is apparent (r = 0.066, p = NS).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

We found that deposition of aerosolized rhDNase I varied widely from patient to patient (the highest 5 times the lowest). The major factors influencing this observation were an age-dependent concentration of particles in the throat and the degree of airway obstruction measured by spirometry.

The younger patients received more deposition in the upper airways. The decrease in throat deposition with increased tidal volume and age may be related to growth and geometric changes in the pharynx and larynx. Patients with more rapid inspiration might deposit more readily in the larynx, but, in this study, patients with higher inspiratory flow had reduced upper airway deposition. If upper airway geometry changes with growth, local flow patterns may be affected (as suggested by changes in deposition), and those patients with increased upper airway deposition may have increased upper airway resistance and therefore, all other things being equal, reduced inspiratory flow.

In an earlier study from our laboratory a similar group of patients with CF inhaling antibiotic aerosols during tidal breathing had significantly less deposition in the upper airways, averaging less than 5% of the total deposited patient dose (17). In the present study, we found an average dose to the upper airways of 15.5% of total deposition (Table 3), with the greatest in the youngest patients. A major difference between the aerosols of the previous study and the present paper was the MMAD; 1.1 µm versus 3.0 µm for rhDNase I. The differences in particle distribution were almost certainly due to the nebulizers and solutions utilized in the different studies. For relatively innocuous drugs, deposition to the pharynx and larynx may be of little significance, but, for rhDNase I, it is now well known that a significant number of patients, using the same recommended nebulizers, report upper airway side effects (hoarseness) (8) and, anecdotally, some will stop therapy.

While deposition varied from patient to patient, standardization of the nebulizer delivery systems minimized variation in drug delivery. For example, in spontaneously breathing patients, treated with aerosolized pentamidine using different unmatched nebulizer systems, inhaled mass varied by over 100% (13). In the present study there were no significant differences in inhaled mass between nebulizers and patient-related differences in airway geometry were the dominant factors affecting lung deposition. This conclusion is based on the observation that lung deposition was most closely correlated with FEV₁. Changes in FEV₁ are a measure of disease progression and our data clearly indicate that the lung dose is greater in the sicker patients. In obstructed patients with chronic obstructive pulmonary disease, changes in airway caliber are known to affect both regional and total lung deposition of aerosols. As FEV₁ declines, sites of deposition shift to central airways (20) and lung deposition increases (21). Therefore, as disease progresses, the concentration of deposited drug tends to increase in the airways. The present study indicates that this process also occurs in CF for 3-µm particles. Our data describe the following overall trend depicted in the images presented in Figure 2; as the patients age, during inspiration, there is a decrease in upper airway deposition and more particles are presented to the lungs. If the patients have more airways disease, lung deposition will be enhanced. Therefore, lung deposition is greatest in those older patients who have more airway obstruction.

Finally, while a dose-dependent effect of aerosolized rhDNase I has been suggested from in vitro measures of mucous transport (5) most clinical studies do not show a consistent effect of increased dosing from 2.5 mg qd to bid (8). After cessation of therapy we found that spirometric decrements in lung function were not closely related to the lung dose of aerosolized rhDNase I suggesting that all patients received an adequate dose from the presently recommended schedule. While it is possible that some effect of lung dose on spirometry might be detected in a larger group of patients, changes in FEV₁ may not be the only indicator or efficacy. To date, we do not have a complete understanding regarding the in vivo mechanisms accounting for the clinical improvements noted with this drug. The decrease in frequency of infections and the sense of well being reported on therapy are other end points that may not be directly related to FEV₁. Factors involving local growth of bacteria and mucociliary and cough clearance may be important and they are the subject of ongoing studies.