Study Design and Subject Populations
This was a double-blind, randomized, placebo-controlled study, consisting of a screening visit and two study visits. The protocol was approved by the Johns Hopkins University Institutional Review Board (IRB). Written informed consent was obtained from parents and either written, or verbal assent, was obtained from the participants depending on their age.
Inclusionary criteria were: males and non-pregnant females, age 7-14 years, diagnosis of CF by sweat chloride > 60 meq/L, or presence of two disease-causing CFTR mutations, and FEV1 ≥ 90% of predicted values.
A group of healthy non-CF adults, who had undergone MCC measurements previously at Johns Hopkins, as part of a multicenter MCC standardization protocol, served as a comparison control group. Clearly, healthy children without CF would have been the preferred control group for these studies. However, our IRB does not allow healthy children to be exposed to ionizing radiation for research purposes.
Children who were taking inhaled tobramycin were studied during the off-month of treatment. Children were asked to stop bronchodilators, rhDNase (dornase-alpha), and airway clearance therapy for 12 hours before and during each study visit.
On the screening visit, children underwent a standardized sputum induction procedure, utilized by the CF Therapeutic Development Network . Children inhaled 3% saline for 12 min using a DeVilbiss Ultra-Neb 99 nebulizer (Sunrise Medical, Somerset, PA). Bacterial and fungal cultures were carried out with any sputum obtained. If sputum was not expectorated, an oropharyngeal culture was obtained. None of the children spontaneously produced sputum at baseline.
Placebo and Hypertonic Saline Administrations
At least one week following the screening visit, children inhaled 5 ml of 0.12% saline (placebo) on one study visit. During a second study visit, they inhaled 5 ml of 7% HS. The order of these two visits was randomized and they occurred at least one week apart. Placebo and HS aerosols were delivered by an LC Star® nebulizer and ProNeb Ultra® compressor (PARI Respiratory Equipment, Richmond, VA). Prior to the administration of HS or placebo, children inhaled 2 puffs of albuterol (Ventolin® CFC, GlaxoSmithKline, London, England) delivered using an Aerochamber® spacer (Trudell Medical International, London, Ontario). Albuterol was administered to minimize the development of bronchospasm and coughing that has been reported following the administration of HS to patients with CF .
We made no attempt to blind the taste of the trial solutions. This was based on advice from our Research Pharmacy, whose staff felt that the addition of any masking agents to the solutions to be aerosolized could have unknown effects on MCC.
Mucociliary Clearance Measurements
Following inhalation of HS or placebo, patients inhaled the radioisotope 99 mtechnetium sulfur-colloid (radioaerosol) and underwent sequential imaging with a gamma camera (ZLC, Siemens, Gammasonics, Des Plains, IL). Radioaerosol was delivered by an LC Plus® nebulizer (PARI) and DeVilbiss Pulmo Aide® compressor (Sunrise Medical). Children inhaled 20-25 times from the nebulizer starting near functional residual capacity at a flow rate of 0.5 L/sec. During inhalation, the nebulizer was pulsed for a fixed time of 0.7 sec. After inhalation to total lung capacity, children exhaled immediately. There was no breath hold.
Lung images were acquired every 2 min for the first 20 min and every 10 min for up to 90 min. Between the 60 and 90 min measurements, children coughed 30 times in two 15 cough sequences. Clearance values reported at 60 min represented removal of radioisotope and mucus from the airways due to MCC, while values at 90 min represented a combination of MCC and cough clearance. We ensured that all study patients coughed with similar effort by monitoring their peak expiratory flow during each cough with a PiKo-1 electronic peak flow meter (Ferraris Medical, Inc., Louisville, CO).
Lung images were analyzed in terms of the amount of radioactivity initially deposited in the right lung at time 0 (i.e. first lung image) and the amount of radioisotope that was retained at each time point thereafter, using a Sopha computer (SMV, Twinsburg, OH). Counts in the right lung were corrected for time decay and background activity and expressed as a percentage of the radioactivity detected at time 0. Percent clearance was calculated as the difference between percent radioactivity in the lung at time 0 and radioactivity remaining at any given time point thereafter.
The central to peripheral (C:P) ratio was determined from the time 0 image using methods that have been described previously . Larger C:P ratios indicated greater deposition of the radioaerosol in the larger, central airways.
Sample Size Estimate
Sample size was estimated using Stata Version 9 (StataCorp, College Station, TX) and was based on MCC data from prior studies in adolescent and young adult patients with CF who were treated with HS  and adult healthy subjects who were treated with placebo . In the former study , the mean difference in MCC at 1 hour was 4.7% (post HS - baseline values). In the latter study , mean MCC ± standard deviation (SD) at 76 minutes was 12.3 ± 8.3 . Since our proposed analysis was a paired comparison of differences within the same child, we assumed a range of values for the SD of the difference in clearance between placebo and HS that was less than the variability of independent measurements observed in these two studies. We determined that a sample size of 12 children would provide 80% power to detect a 5% or larger mean difference in MCC between inhalation of placebo versus HS aerosol, assuming the SD of the paired differences was 6% and assuming a = 0.05 (2-sided). A 5% mean difference between treatments is similar to what was reported by Donaldson et al. .
Because of the small subject population, non-parametric analyses were used throughout. Data are presented as box plots showing the median and interquartile range (IQR). The "whiskers" represent the 10th and 90th percentiles. Clearance values are expressed as percent MCC at 60 min (MCC60) and at 90 min (MCC90) after inhalation of the radioaerosol. MCC and C:P ratios for the children on the placebo visit and for the adults were compared using a Wilcoxon rank sum test. The change in MCC60 and MCC90 with HS from placebo in the children was compared using a Wilcoxon signed rank test. In summarizing the results, it was noticed that some children had lower percent MCC values on the placebo visit than others. A post hoc subgroup analysis on the change in MCC60 and MCC90 with HS from placebo was conducted using a median split of MCC60 and MCC90 on placebo to compare children with lower MCC to higher MCC using a Wilcoxon rank sum test. C:P ratios in these subgroups were also compared using a Wilcoxon rank sum test. Non-parametric tests were conducted using SAS version 9.22 (SAS Institute, Inc., Cary, North Carolina). All tests were 2-sided and significance was set at P < 0.05.