Physiologic evidence for the efficacy of positive expiratory pressure as an airway clearance technique in patients with cystic fibrosis.

Cystic fibrosis (CF) is the most common lethal genetic disease affecting Caucasians. (1) Increased airway secretions and subsequent bacterial lung infections result in the development of chronic obstructive pulmonary disease (COPD) in individuals with CF. (2) Complications of the lung disease are responsible for increased mortality and morbidity. (1) Accumulation of airway secretions leads to airway inflammation predisposing to airway narrowing, increases in airflow resistance, (3,4) and gas trapping. (5,6) This results in non-uniform distribution of ventilation (7,8) and impaired gas mixing. (3,7) These changes ultimately are manifested as reductions in gas exchange that may limit functional capacity.

Effective airway clearance is a critical component of the management of CF. The most commonly used airway clearance treatment (ACT) is often called "chest physical therapy" (CPT) and includes gravity-assisted postural drainage and manual percussion and vibration techniques. These interventions are time and effort consuming and create dependency on others, which may contribute to poor adherence to airway clearance in these patients. (9) Because individuals with CF are surviving longer and leading independent lifestyles, there is a greater need for ACTs that do not require assistance. (10)

Alternative ACTs, including positive expiratory pressure (PEP) breathing, have been introduced in an effort to provide effective secretion clearance while promoting treatment adherence, fostering patient independence, and minimizing physical discomfort. (10) Positive expiratory pressure breathing is somewhat similar to pursed-lip breathing in that a resistance to expiration is applied at the mouth during expiration. This results in increased pressure at the mouth that is transmitted to the airways and acts to hold the airways open during expiration. (11) The increased airway pressure during expiration is thought to prevent premature airway closure and thus reduce gas trapping in the lung. (11) In addition to holding the airways open and prolonging expiratory airflow, PEP is purported to promote movement of mucus proximally. (12) Positive expiratory pressure breathing has been shown to be effective in secretion removal for patients with CF. (13-15) Additionally, patient-administered PEP breathing is well accepted by patients, (14-16) easy to use, (14-16) time efficient, (14,16) and inexpensive (14) compared with postural drainage, manual percussion, and vibration.

Positive expiratory pressure is produced by breathing through a face mask fitted with an expiratory resistance and a pressure manometer (Fig. 1). There are 2 forms of PEP: low PEP and high PEP. Low PEP involves tidal volume inspirations and slightly active expirations against resistances that produce pressures at the mouth of 10 to 20 cm [H.sub.2]O during exhalation. (14,16) The high-PEP technique uses high lung volumes and forced expiratory maneuvers against resistances that generate expiratory pressures greater than 20 cm [H.sub.2]O. (15) The amount of resistance used in high PEP is determined for each patient using spirometric procedures. (15) The target resistance in high PEP is the one that generates a pressure that allows the patient to produce a forced vital capacity (FVC) that is greater than the FVC produced with no PEP. The increase in FVC, or volume of gas that can be forcefully and rapidly expired after a maximal inspiration, with high PEP indicates that additional residual volume (RV) gas, or gas remaining in the lungs at the end of a maximal expiration, was evacuated from the lung before airway closure occurred. (5)

[FIGURE 1 OMITTED]

In addition to enhancing airway secretion clearance, PEP breathing treatments are effective at improving lung volumes (15,16) and expiratory flow (15,16) when used over a 10-month (15) to 12-month (16) period. Reports on 40 patients with CF who experienced similar pulmonary dysfunction at study entry and who were randomly assigned to either a low-PEP group or a CPT group indicated that changes in FVC and forced expiratory volume in 1 second ([FEV.sub.1]), a measure of larger, central airway expiratory flow, were greater following 1 year of low PEP compared with 1 year of the use of interventions associated with CPT. (16) In the same study, however, McIlwaine and coworkers (16) found no differences between low PEP and CPT when preintervention to postintervention changes were compared for forced expiratory flow during the middle half of the FVC maneuver ([FEF.sub.25%-75%]), a measure of smaller, peripheral airway expiratory flow. Similarly, most reports of small, yet measurable, improvements in lung function generally showed no differences after low PEP compared with after traditional interventions for a single treatment (14,17) or after multiple treatments. (13)

Beneficial effects on lung function have been documented following the use of high PEP in individuals with CF. Oberwaldner et al (15) reported that 10 months of high PEP resulted in improvements in lung function and higher mean daily sputum volumes cleared compared with 2 months of CPT interventions. High-PEP breathing (mean pressures of 61 cm [H.sub.2]O, range = 26-102) resulted in reduced gas trapping, increased expiratory flow, and improved lung volumes over a 10-month period compared with CPT, (15) suggesting that the pressures used in PEP treatments may need to be relatively high to allow airways to be effectively held open and trapped gas and mucus to be evacuated.

Although the low-PEP technique (13,14,16,17) is more widely used than high PEP, (15) there is a lack of scientific evidence to support this practice. An examination of the effects of low and high PEP within the same individuals has not been made. Such an investigation may provide insight into the effects of the technique on airflow mechanics and help to identify the underlying physiologic mechanisms for effective airway mucus removal. Describing the physiologic changes following PEP would add to the body of evidence required to assess its utility and provide guidelines for the efficacious application of these techniques.

A putative goal of any airway clearance technique is to decrease airway obstruction and airway resistance and improve distribution of ventilation through the mobilization and removal of secretions. (18) Although several researchers have evaluated the effects of PEP on lung function (12,14-16) and sputum removal, (13-15) none have reported the effects of PEP breathing on distribution of ventilation and gas mixing.

The primary purpose of our investigation was to describe responses in distribution of ventilation, gas mixing, lung volumes, and expiratory airflow following low-PEP and high-PEP treatments for patients with moderate to severe CF lung disease. A secondary purpose of this investigation was to describe responses in percentage of arterial blood oxyhemoglobin saturation ([Spo.sub.2]) and the amount of mucus expectorated following low-PEP and high-PEP treatments. In our experience, deep breathing and coughing performed during lung function testing stimulates airway mucus loosening and expectoration. We, therefore, included a no-PEP treatment session to describe the responses following pulmonary function testing alone on these physiological measures.

Theory of Single-Breath Inert Gas Test Measurement of Ventilation Distribution

In people without known pathology or impairments, under normal physiologic conditions, there are interregional differences in the distribution of ventilation such that each breath of inspired air is unevenly distributed within the lungs. (19) The lung base receives the greatest portion of the inspired breath, while the lung apex receives the smallest portion of the same inspired breath. (19) Intraregional distribution of ventilation, however, is nearly uniform for healthy lungs. In diseased lungs, distribution of ventilation may be non-uniform. Some lung units will be well ventilated, whereas other lung units, within the same region, will be poorly ventilated. (20) The uniformity of the distribution of ventilation can be measured using a single-breath inert gas test. (20) In a single-breath inert gas test, a gas mixture containing an inert test gas, such as helium, is slowly and evenly inhaled from RV to total lung capacity (TLC) and, without pausing, exhaled slowly to RV again. The foreign test gas marker, helium, is diluted by gas already residing in the lung, and the exhaled helium concentrations are sampled continuously at the mouth and analyzed.

Figure 2 shows the distribution of ventilation depicted on a single-breath inert gas test curve of an exhaled inert gas concentration plotted against exhaled lung volume following inhalation of a gas mixture containing the test gas marker 9% helium. The single-breath curve consists of 4 phases (Fig. 2). At the start of expiration, which is phase I, the air that is initially exhaled contains 9% helium because this air comes from the dead space of the upper airways where no gas exchange is occurring. Phase II gases are from the dead space and alveolar gas (the helium has mixed with RV gas at the alveoli), and therefore the helium concentration begins to decrease. Phase III is the alveolar gas plateau phase, which is horizontal to the x-axis when the expired concentration becomes consistent, indicating that equal amounts of helium are being emptied from all lung units and distribution of ventilation is homogeneous. A continuous decrease in expired helium concentration or a downward phase III slope, away from horizontal, indicates asynchronous filling and emptying among lung units because of alterations in lung tissue distensibility and increases in airway resistance, (20,21) which lead to ventilation inhomogeneity. (20,21) Airway obstruction secondary to pulmonary secretions contributes to airway resistance. Phase III in individuals with airway obstruction, who have nonuniformly distributed test gas because of asynchronous lung unit filling and emptying, will display a continuous fall in expired helium gas concentration, away from horizontal. Phase IV marks the abrupt onset of a decrease in the helium concentration as small, helium-rich airways close. In patients with obstructive airway disease, phase IV onset occurs early compared with individuals with healthy lungs due to small airway closure.

[FIGURE 2 OMITTED]

Theory of Dilution Index

Single-breath curve data were standardized by transforming absolute expired helium gas concentration data to dilution index (DI) format. The DI was defined by Fowler (20) as the ratio of an added inspired volume of gas to the original volume of gas already residing in the lungs during the single-breath washout test for distribution of ventilation.

The greater the expired concentration of marker gas, the greater the DI value (Fig. 3). Good gas mixing results in a high DI value such as 5 and is observed when a relatively large inspired volume of gas mixture containing test gas does not become very diluted when added to a relatively small RV of gas already present in the lung. By comparison, a low DI value (eg, 1.5) indicates extensive dilution of a small, inspired volume containing test gas by a relatively large RV, as is often the case for older individuals or those with COPD. (20,21)

[FIGURE 3 OMITTED]

Methods

Subjects Six subjects with CF documented by sweat test (22) were enrolled in the study. Experimental procedures were explained to 23 patients with CF from a pool of 120 patients from the Children's Lung and Cystic Fibrosis Center, Children's Hospital of Buffalo, who met the study inclusion criteria. Large geographical distances patients had to travel, lack of patient availability on weekends when data collection occurred, and patient participation in concurrent studies precluded participation in this study by many patients followed at the Children's Lung and Cystic Fibrosis Center, and, therefore, active recruitment for the current study was limited to 23 patients. Seventeen of the 23 patients who met study inclusion criteria chose not to participate due to the high time commitment required by the study protocol.

Subjects were eligible if they were medically stable and had not been hospitalized during the previous month for management of a pulmonary exacerbation. Patients had to be able to perform lung function testing according to American Thoracic Society (ATS) criteria (23) and must not have missed more than 2 scheduled clinic or research study appointments within the previous year. Study participants could not be on supplemental oxygen, have a history of pneumothorax, or perform PEP breathing routinely. Individuals who were on oxygen were excluded because the rigorous demands of the protocol raised concerns about their ability to complete the study. Individuals who had a history of pneumothorax were excluded for safety reasons related to breathing against high resistances. We anticipated that the effects of PEP breathing on airway mechanics could potentially be different for individuals who routinely used PEP as compared with those who did not use PEP routinely. Only one patient who met all other inclusion criteria was deemed ineligible to participate in the study because he performed low PEP on a regular basis for the purpose of airway clearance.

Prior to participation, informed consent was obtained from all study volunteers and parents (for subjects younger than 18 years of age). Testing was performed on 6 subjects (3 male, 3 female), aged 13 to 22 years, with CF. The data of one subject were excluded when it was determined that there was an exacerbation of his chronic lung infection.

Subject characteristics at the time the study began are presented in Table 1. Lung function test results indicated patients overall had moderate to severe pulmonary dysfunction according to ATS criteria (24) (Tab. 1). Subjects had moderate central airways obstruction (24) and severe peripheral airflow limitation, (24) as indicated by percentage of predicted FE[V.sub.1] and [FEF.sub.25%-75%], respectively (25,26) (Tab. 1). Ventilation distribution and gas mixing were disturbed in the subjects with CF, as indicated by baseline values for the slope of the line representing DI plotted against expired lung volume ([S.sub.D1]/volume]) and low gas mixing (DI) values.

Experimental Protocol

Each subject visited the Pulmonary Function Laboratory at the Children's Lung and Cystic Fibrosis Center on 3 separate days. On day 1, we collected data before and after lung function testing. Low- and high-PEP breathing were randomly assigned to days 2 and 3, respectively. Days 2 and 3 were always separated by 5 days in order to allow for the washout of any carryover effects of PEP intervention. Single-breath inert gas studies and lung function tests were performed prior to (test 1), immediately following (test 2), and 45 minutes following (test 3) 20 minutes of quiet sitting, low-PEP breathing, or high-PEP breathing, as shown in Figure 4. Sputum was collected during each test (test 1, test 2, and test 3), treatment, and rest period. Sputum was collected during the tests because of the effectS of deep breathing and coughing performed during lung function testing on mucus loosening and removal. Arterial blood oxyhemoglobin saturation was estimated using continuous, noninvasive monitoring of finger pulse oximetry (Nellcor Sympony N-3000)* at baseline, following each test and rest period, and during all intervention conditions.

[FIGURE 4 OMITTED]

No PEP breathing (day 1). Subjects sat quietly for 20 minutes during the no-PEP breathing period. Subjects were encouraged to cough and to clear secretions every 4 minutes.

PEP breathing (days 2 and 3). The PEP breathing system consisted of an anesthesiology face mask fitted with one-way inspiratory and expiratory valves (Fig. 1). expiratory resistors with internal diameters ranging between 1.5 and 5.0 mm were connected to the expiratory port. A pressure manometer was interfaced with the resistor to provide visual feedback so that correct PEP measurements could be maintained. expiratory resistors and mean sustained expiratory pressures generated during low-PEP and high-PEP breathing, for each subject, are shown in Table 2.

For low-PEP breathing, a resistor with an internal diameter that gave a steady PEP of 10 to 20 cm [H.sub.2]O during expiration while breathing through the PEP mask was used. (14) A steady PEP of 10 to 20 cm [H.sub.2]O was maintained during tidal exhalations, and exhalations were slightly active. (14,16) Subjects breathed against the expiratory resistance for 8 to 10 breaths, then they came off the PEP mask and were encouraged to cough as much as necessary to clear secretions following each cycle of 8 to 10 breaths. Six cycles of this procedure were performed.

For high-PEP breathing, the internal diameter of the expiratory resistor to be used was determined according to previously described procedures. (15) The outlet valve of the mask was connected to the intake port of the spirometer, and subjects then performed a minimum of 3 FVC maneuvers through each of the 8 resistors. Subjects progressed through increasing resistance settings. The resistor at which "FVC with PEP" exceeded "FVC with no PEP" was selected for PEP breathing. In the case where more than one resistor produced an "FVC with PEP" that exceeded an "FVC with no PEP," the lowest resistance was selected. Onset of fatigue was defined as a stepwise decline in expiratory pressures with each breath or the inability to sustain consistent expiratory pressures throughout the 8 to 10 breaths.

During high-PEP breathing, subjects were prompted to breathe against the selected resistance for 8 to 10 breaths, inspiring a volume of air larger than a normal tidal volume breath, with active contraction of the abdominal muscles during exhalation. (15) Following each cycle of 8 to 10 breaths, subjects were prompted to inhale to TLC and perform a forced expiratory maneuver into the mask against the resistor. (15) During the forced expiratory maneuver to a low lung volume, secretions were usually mobilized and coughing was stimulated. Six cycles consisting of 8 to 10 breaths, followed by a forced expiratory maneuver and coughing, were performed. Each subject was allowed to determine how much coughing was necessary in order to clear secretions during and following each cycle of PEP breathing.

Measurements

Single-breath inert gas test. A modification (27) of Fowler's single-breath nitrogen test (20) was used to assess distribution of ventilation and gas mixing. The breathing circuit for the single-breath test (28) consisted of a mouthpiece-valve system and an electronic spirometer ([dagger]) interfaced with a bag-in box system (Ohio 840). ([double dagger]) Gas was analyzed continuously at the mouthpiece by a capillary line connected to a mass spectrometer gas analyzer (MGA-1100), ([section]) interfaced with a computer and the Spike II software program. ([parallel]) Changes in volume and flow were measured by electrical outputs from the 10-L spirometer that was interfaced with the computer software. Subjects were prompted to breathe a single breath of gas mixture containing 9% helium, 21% oxygen, and 70% nitrogen. The goal was to obtain 3 acceptable helium gas concentration versus volume craves per subject at each measurement interval. (29,30)

Spike II records of time, flow, volume, and absolute helium concentration data were later transferred to Microsoft Excel software. # For each single-breath test, absolute expired helium gas concentrations were normalized by expressing the data as a percentage of the mean inspired helium gas concentration value for a particular single-breath test in order to control for variability in the expired helium gas concentrations. (20,21) Normalized expired helium gas concentrations were then expressed in DI format. (20,21)

Distribution of ventilation (phase III slope data expressed as [S.sub.DI]/volume). A regression test was performed, using Microsoft Excel software, on the DI versus lung volume data between the onset of phases III and IV of the single-breath curve (Fig. 2). The slope of the regression line represented the DI/volume slope of phase III, a measure of uniformity of the distribution of ventilation. (20,21) The closer the phase III slope was to horizontal, the more uniform the distribution of ventilation. A downward slope of the line away from horizontal represented an increase in the phase III slope and poorer distribution of ventilation.

Gas mixing (DI values expressed at an absolute lung volume [[DI.sub.VL]]). In addition to the slope data, we also analyzed DI data at an absolute lung volume ([DI.sub.VL]) to determine if there was a change in expired helium concentrations between test intervals without there being a change in DI/volume slope values. (21) The DI values were identified at 50% of the preintervention expired vital capacity. The DI values were obtained immediately after and 45 minutes after intervention at the same absolute lung volume that was noted at 50% of preintervention expired lung volume for that day. This procedure was performed for each intervention condition for each subject.

The [S.sub.DI/volume] and [DI.sub.VL] measurements for each subject were obtained from the average of 2 or 3 single-breath curves at each time point (baseline, immediately after intervention, and 45 minutes after intervention) for each intervention condition.

Lung function tests. Lung function was assessed by the same investigator, who was extensively trained in pulmonary function testing. Vital capacity and expiratory flow measurements were obtained by simple spirometry, and lung volumes were measured using body plethysmography (MedGraphics Model 1070, Series 2) ** interfaced with a MedGraphics Breeze software program.

Flow-volume curves were generated from the best of 3 forced expiratory maneuvers to assess FVC, [FEV.sub.1], and [FEF.sub.25%-75%] according to ATS standardized guidelines. (23) A constant volume body plethysmograph was used to determine thoracic gas volume ([V.sub.TG]), from which RV was calculated. (31,32) The largest acceptable slow vital capacity (SVC) value was saved and used for calculation of RV according to the methods of Dubois and coworkers. (31,32) We reported SVC instead of FVC because the single-breath test involves a slow maneuver that minimizes dynamic compression and airway collapse, which are associated with forced expiratory maneuvers. The SVC and RV measurements were made because PEP breathing can lead to changes in lung volume. (15) Specifically, changes in RV, or the original volume, and changes in SVC, or the inspired volume, may potentially lead to changes in gas mixing. (21) Lung function data were expressed as percentages of predicted values. (26,31,33)

Single-breath and lung function test measurements were made according to standardized procedures. (23,29,31,32) Single-breath inspiratory and expiratory vital capacities, for the same maneuver, were required to be within 5% of one another to accept any test. (29) Expired vital capacities between trials needed to be within 5% for each individual. (29) Lung function measurements were repeated 3 times or until values were within 5% of one another. (23,31,32) The covariance of these measures is consistent with previously published work. (34-36)

Sputum dry weight. Expectorated sputum was collected into preweighed dry specimen cups. Wet and dry (following 4 days of drying in an oven) sputum weights were recorded.

Data Presentation

Group means were calculated for demographic data, DI/volume slope, [DI.sub.VL], and [Spo.sub.2] and for percentage of predicted [FEV.sub.1], [FEV.sub.1]/FVC, [FEF.sub.25%-75%], SVC, and RV at entry to the study. Group data were calculated and presented as means for DI/volume slope, [DI.sub.VL], SVC, RV, [FEV.sub.1], [FEF.sub.25%-75%], [Spo.sub.2], and sputum dry weight at each measurement interval during all conditions.

Results

Distribution of Ventilation ([S.sub.DI/volume])

Distribution of ventilation worsened, as assessed by the phase III alveolar slope of the single-breath inert gas test, immediately after intervention (test 2) and 45 minutes after intervention (test 3) as compared with before intervention (test 1) for the low-PEP and high-PEP groups (Fig. 5). During the low-PEP protocol, the downward slope in phase III increased, on average, 25% immediately after intervention and 35% 45 minutes after intervention from the preintervention slope. During high PEP breathing, the downward slope in phase III increased, on average, 24% immediately after intervention and 39% 45 minutes after intervention from the preintervention slope. The downward slope in phase III increased, on average, 3% immediately following no PEP and decreased 1% 45 minutes after intervention.

[FIGURE 5 OMITTED]

Gas Mixing (]DI.sub.VL])

Gas mixing within the lung improved in all groups across time such that, by 45 minutes after intervention, the no-PEP group had improved, on average, 5%, the low-PEP group had improved, on average, 15%, and the high-PEP group had improved, on average, 23% (Fig. 6).

[FIGURE 6 OMITTED]

Percentage of Predicted Slow Vital Capacity (SVC % Predicted)

Slow vital capacity increased, on average, 5% in both the low-PEP and high-PEP groups immediately after intervention (Fig. 7). A 9% average increase in SVC occurred in the low-PEP group by 45 minutes after intervention and the high-PEP group showed a 13% average increase in SVC by 45 minutes after intervention as compared with before intervention. Slow vital capacity, increased by 1% immediately following intervention and remained unchanged 45 minutes following intervention in the no-PEP group.

[FIGURE 7 OMITTED]

Percentage of Predicted Residual Volume (RV % Predicted)

Residual volume decreased, on average, 4% immediately following low-PEP breathing and 20% 45 minutes following low-PEP breathing as compared with before intervention. During high-PEP breathing, RV decreased, on average, 10% immediately after intervention and 30% 45 minutes after intervention as compared with before intervention. Residual volume increased, on average, 5% immediately following no-PEP breathing followed by a 13% average decrease at 45 minutes after intervention as compared with before intervention (Fig. 8).

[FIGURE 8 OMITTED]

Expiratory Airflow ([FEV.sub.1] and [FEF.sub.25%-75%])

Forced expiratory volume in 1 second increased, on average, 6% immediately following low-PEP breathing and 7% 45 minutes following low-PEP breathing. During the high-PEP condition, FE[V.sub.1] increased, on average, 5% immediately following intervention and 9% 45 minutes following intervention. There was less than a 1% increase in FE[V.sub.1] following the no-PEP condition and less than a 1% decrease in [FEV.sub.1] at 45 minutes following the no-PEP condition. During the low-PEP condition, FE[F.sub.25%-73%] increased, on average, 29% immediately "after intervention and 1% 45 minutes "after intervention compared with before intervention (Tab. 3). A 27% increase in FE[F.sub.25%-75%] occurred immediately following high-PEP breathing followed by a 22% increase 45 minutes following high-PEP breathing as compared with before intervention. The FE[F.sub.25%-75%] increased, on average, 2% immediately following the no-PEP condition and 7% 45 minutes following the no-PEP condition (Tab. 3).

Percentage of Arterial Blood Oxyhemoglobin Saturation Percentage of [Spo.sub.2] increased approximately 1% following the initial lung function testing and remained at this level immediately following the low-PEP and high-PEP interventions, whereas the no-PEP intervention resulted in percentage of [Spo.sub.2] returning to preintervention levels. The second lung function test was associated with reductions in percentage of [Spo.sub.2] of approximately 1% in the low-PEP and high-PEP conditions, whereas the no-PEP condition was associated with no changes. Following the 45-minute postintervention period, percentage of [Spo.sub.2] continued to decline in the low-PEP condition (~1%), whereas modest increases and decreases (<1%) in percentage of [Spo.sub.2] were observed in the high-PEP and no-PEP conditions, respectively. The percentage of [Spo.sub.2] increased with low-PEP breathing (1.5%) and with high-PEP breathing (1%) following the final lung function test (Fig. 9).

[FIGURE 9 OMITTED]

Sputum Dry Weight

Cumulative dry weights for sputum increased across time in all groups (Fig. 10). An increase in sputum expectoration occurred initially, after the first lung function test period, and continued to increase throughout the remainder of the protocol. The largest increases in sputum expectoration were in response to lung function testing as compared with during, immediately after, and 45 minutes after intervention.

[FIGURE 10 OMITTED]

Discussion

The main finding of our study was that gas mixing improved for all intervention conditions and particularly for the low-PEP and high-PEP conditions. Twenty minutes of PEP breathing was effective at improving gas mixing in a small group of medically stable patients with moderate to severe CF-related lung disease. The increases in expired helium gas concentrations, particularly after high-PEP breathing and to a lesser extent following low-PEP breathing, we believe, reflected the improved efficiency with which an inspired gas mixture containing test gas helium mixed with gas already residing in the lung (28) after intervention. This finding indicates that, during PEP breathing, previously closed airways were opened, additional residual volume gas was exhaled, and a larger inspired volume of gas entered the lung, potentially improving gas exchange. Any improvements in gas mixing following the no-PEP condition most likely would have been due to airway mucus loosening and removal stimulated by deep breathing and coughing performed during lung function testing. Percentage of [Spo.sub.2] increased during the high-PEP protocol. In addition, there was a cumulative increase in the amount of sputum expectorated during the 3 intervention conditions.

Gas-Mixing ([DI.sub.VL]

The improvements ill gas mixing following the low-PEP and high-PEP conditions might be explained by the effects of 20 minutes of repeated and prolonged exhalations against PEP resistance on time constants of lung units. A lung unit is the functional gas exchange unit of the lung and consists of alveolated structures distal to the end of the terminal bronchiole. The time constant for a lung unit is defined as the time it takes a lung unit to empty or fill and is equal to the product of its resistance (R) to airflow and its compliance (C), R (cm [H.sub.2]O/L/s) x C (L/cm [H.sub.2]O), making the movement of air dependent on airway diameter and tissue elasticity,. Time constants are slow when lung units have low distensibility and high airway resistance such as in CF-related COPD. Parallel lung units, residing in the same lung region, that are exposed to the same inflation and deflation pressures do not uniformly fill and empty in the presence of obstructive lung disease as compared with parallel lung units in healthy lungs that have nearly the same filling and emptying times. (8) The low expired helium concentrations measured before intervention in our study reflected heterogeneity of time constants within the peripheral airways of the subjects with CF, confirming the presence of lung units with fast and slow time constants, which is consistent with previous reports. (3,8) We believe the small improvements in gas mixing following the no-PEP condition were likely due to deep breathing and coughing, which facilitated sputum mobilization and removal and reduced airway obstruction.

In contrast to the lack of homogeneity prior to intervention, it is likely that homogeneity among time constants was increased during PEP breathing. Resistance breathing dilates peripheral airways and facilitates the ongoing exhalation of RV gas. (15) The continual exhalation of RV gas generates airflow through smaller airways and purportedly mobilizes airway mucus in these areas. (15) Less peripheral airway obstruction means faster filling and emptying times for all lung units but particularly for slow lung units? Gas mixing improved because time constants for lung units became faster, thereby augmenting the exhaled gas volume during and following PEP (Fig. 6).

Airflow generated by the continual exhalation of RV gas may support our preliminary finding that mucus can be mobilized and expectorated during and following 20 minutes of low-PEP and high-PEP breathing. We also observed that high-PEP breathing was of particular benefit, leading to an increase in the amount of sputum expectorated. Our subjects reported less chest unpleasantness due to pulmonary secretions following low-PEP and high-PEP breathing. Three subjects reported greater ease of breathing and less chest unpleasantness following high PEP breathing. Positive expiratory pressure breathing, in general, and high PEP, in particular, appeared to alter gas mixing positively and enhance airway clearance. We believe deep breathing and coughing and the prolonged, forceful exhalations performed during lung function testing facilitated sputum mobilization and removal as shown in Figure 10 and, therefore, contributed to the cumulative increase in sputum amounts during the 3 intervention protocols. We believe the improvements in percentage of [Spo.sub.2] (Fig. 9) during the high-PEP protocol can be attributed to the improvements in both gas mixing and sputum removal.

Lung Volumes

Other researchers have reported improvements in FVC (14-16) and RV (15) following PEP breathing but without investigation of the effects of these improvements in lung volumes on gas mixing. Our data suggest that gas mixing and SVC improvements were likely due to a reduction in the complete or partial obstruction of peripheral airways.

In the presence of severe peripheral airflow limitation, as indicated by the low [FEF.sub.25%-75%] in our study, early airway closure and dynamic compression of smaller peripheral airways occurs during active exhalation at rest, (15) as reflected by the low SVC values and the high RV values (Tab. 1). But, when low PEP is applied at the airway opening during an active exhalation, PEP breathing increases luminal pressure, thereby keeping airways open. (11,15) The pressure drop down the airway is slower during PEP breathing as compared with breathing without PEP. (15) As exhalation progresses toward RV, the luminal pressure drops slowly in the direction of the airway opening. (11-15) The slow drop in luminal pressure, while exhaling against PEP resistance, prevents early collapse of smaller, peripheral airways so that additional gas volume is exhaled. (15) The additional expired gas volume led to a reduced RV and an increased SVC. The steady decreases in RV following PEP breathing found in our study and observed by other researchers (15) suggest that resistance breathing has a dilating effect on the airways. Breathing at high lung volumes, exhaling against high resistance to RV prior to coughing, and initiating coughing at low lung volume may have delayed the onset of airway closure and prolonged expiratory airflow during high PEP and may have accounted for the enhanced mucus removal during the high-PEP breathing protocol.

Expiratory Airflow

Reports of other researchers (15) coupled with our own observations that [FEV.sub.1], a measure of central expiratory airflow, and [FEF.sub.25%-75%], a measure of peripheral expiratory airflow, improved following PEP breathing supports the concept that PEP breathing facilitates mucus removal through augmenting airflow mechanics. The immediate changes following both levels of PEP breathing in central expiratory airflow were sustained 45 minutes following the interventions, whereas changes in peripheral expiratory airflow were not sustained. The smaller peripheral airways in individuals with CF are unstable and are easily compressed and collapsed, thereby impeding expiratory airflow even during tidal breathing while at rest. (3,4) Long-term high-PEP breathing has been shown to improve peripheral airway function, (15) yet long-term low-PEP breathing was not shown to be as effective at improving [FEF.sub.25%-75%.] (16)

Ventilation Distribution ([S.sub.DI/volume])

The downward phase III slopes for DI plotted against expired lung volume, at baseline, indicated asynchronous filling and emptying by lung units for the subjects with CF. One factor likely contributed to the lack of improvement and the worsening of ventilation distribution, reflected in [S.sub.DI/volume] following either the low-PEP or high-PEP condition (Fig. 5). Airways most likely returned to their "pre-PEP" resting positions immediately following the brief 20-minute period of PEP breathing. Improvements in ventilation and reductions in trapped gas have been detected by other researchers (11) only while airways were stented with a low-pressure resistor in-line with pulmonary function testing equipment following 15 minutes of tidal breathing against a steady low PEP of 10 to 20 cm [H.sub.2]O. In our study, worsening of ventilation distribution after PEP breathing suggests that helium gas likely diffused into previously closed, yet partially obstructed, airways. We cannot, however, ignore the fact that the diffusivity of helium, in part, could explain why there was not a reduction in phase III slopes following PEP breathing. (37) Helium gas may have readily diffused into poorly ventilated regions before PEP breathing and, therefore, minimized our ability to measure changes in the uniformity of ventilation distribution that occurred as a result of PEP breathing. Mobilized mucus also may have had a temporary worsening affect on lung unit filling and emptying times before any improvements associated with mucus removal could be observed.

Conclusion

Limitations of our study include the small sample size and the lack of control for cough frequency during the 3 intervention conditions. We believe, however, that the gas mixing findings are physiologically meaningful and warrant further investigation, especially because there were also changes in [Spo.sub.2]. Although our subjects were encouraged to cough at specified intervals during all conditions, we observed that their adherence was not good.

We recommend that patients with moderate to severe pulmonary dysfunction associated with their CF who use PEP breathing be monitored on a case-by-case basis for improvements or decline in (1) gas mixing, (2) ventilation distribution, (3) lung volumes, (4) expiratory airflow, (5) sputum removal, and (6) [Spo.sub.2] in order to provide evidence for the continued therapeutic use of either level of PEP breathing. Accumulation of evidence-based treatment data will be important to clinicians in making future decisions regarding PEP breathing for patients with CF.

We described the physiologic changes following PEP breathing therapy. This study is the first attempt, to our knowledge, to examine the physiologic basis of low-PEP and high-PEP breathing for the same subjects. Improvements in gas mixing were likely due to augmentation of airway mechanics, which led to improvements in lung volumes, expiratory airflows, sputum removal, and [Spo.sub.2].

All authors provided concept/idea/research design and writing. Dr Darbee and Dr Cerny provided data analysis and fund procurement. Dr Darbee provided data collection and project management, and Dr Cerny provided institutional liaisons. The authors thank Drucy Borowitz, MD, Director, Children's Lung and Cystic Fibrosis Center, Children's Hospital of Buffalo, for providing subjects and facilities/equipment.

The study was approved by the Health Related Professions Human Subjects Review Board, State University of New York at Buffalo, and the Institutional Review Board of the Children's Hospital of Buffalo.

This study was funded by grants from the Foundation for Physical Therapy.

* Mallinckrodt Inc, a company of Nellcor-Puritan Bonnet Co, 675 McDonnel Bird, Hazelwood, MO 63042.

([dagger]) Warren E Collins Inc, 220 Wood Rd, Braintree, MA 02184.

([double dagger]) Ohio Medical Products, Houston, TX 77030.

([section]) Perkin-Elmer Medical Products, 2771 N Garey Ave, Pomona, CA 91767.

([parallel]) Cambridge Electronic Design Ltd, Science Park, Milton Road, Cambridge, United Kingdom CB4 0FE.

# Microsoft Corp, One Microsoft Way, Redmond, WA 98502.

** Medical Graphics Corp, 350 Oak Grove Pkwy, St Paul, MN 55127.

References

(1) Iannuzzi MC. Cystic fibrosis: genetics. In: Davis PB, ed. Cystic Fibrosis. New York, NY: Marcel Dekker Inc; 1993:1-27.

(2) Welsh MJ, Ramsey BW. Research on cystic fibrosis: a journey from the heart house. Am J Respir Crit Care Med. 1998;157:148-154.

(3) Anderson PJ, Blanchard JD, Brain JD, et al. Effect of cystic fibrosis on inhaled aerosol boluses. Am Rev Respir Dis. 1989;140:1317-1324.

(4) Lamarre A, Reilly BJ, Bryan AC, Levison H. Early detection of lung function abnormalities in cystic fibrosis. Pediatrics. 1972;50:291-297.

(5) Desmond KJ, Coates AL, Martin JG, Beaudry PH. Trapped gas and airflow limitation in children with cystic fibrosis and asthma. Pediatr Pulmonol. 1986;2:128-134.

(6) Featherby EA, Weng AT, Crozier DN, et al. Dynamic and static lung volumes, blood gas tensions, and diffusing capacity in patients with cystic fibrosis. Am Rev Respir Dis. 1970;102:737-749.

(7) DeMuth GR, Lowatt WF, Talner NS. Intrapulmonary gas distribution in cystic fibrosis. Am J Dis Child. 1962; 103:52-59.

(8) Brown JS, Gerrity TR, Bennett WD. Effect of ventilation distribution on aerosol bolus dispersion and recovery. J Appl Physiol. 1998;85: 2112-2117.

(9) Passero M, Remor B, Salomon J. Patient reported compliance with cystic fibrosis therapy. Clin Pediatr (Phila). 1981;20:264-268.

(10) Oermann CM, Swank PR, Sockrider MM. Validation of an instrument measuring patient satisfaction with chest physical therapy techniques in cystic fibrosis. Chest. 2000;118:92-97.

(11) Groth S, Stafanger G, Dirksen H, et al. Positive expiratory pressure (PEP-Mask) physiotherapy improves ventilation and reduces volume of trapped gas in cystic fibrosis. Bull Eur Physiopathol Respir. 1985;21: 339-343.

(12) Andersen J, Qvist J, Kann T. Recruiting collapsed lung through collateral channels with positive expiratory pressure. Scand J Respir Dis. 1979;60:260-266.

(13) Braggion C, Cappelletti LM, Cornacchia M, et al. Short-term effects of three chest physiotherapy regimens in patients hospitalized for pulmonary exacerbations of cystic fibrosis: a crossover randomized study. Pediatr Pulmonol. 1995;19:16-22.

(14) Falk M, Kelstrup M, Andersen JB, et al. Improving the ketchup bottle method with positive expiratory pressure, PEP, in cystic fibrosis. Eur J Respir Dis. 1984;65:423-432.

(15) Oberwaldner B, Evans J, Zach M. Forced expirations against a variable resistance: a new chest physiotherapy method in cystic fibrosis. Pediatr Pulmonol. 1986;2:358-367.

(16) McIlwaine PM, Wong LT, Peacock D, Davidson AGF. Long-term comparative trial of conventional postural drainage and percussion versus positive expiratory pressure physiotherapy in the treatment of cystic fibrosis. J Pediatr. 1997; 131:570-574.

(17) Hofmeyr JL, Webber BA, Hodson M. Evaluation of positive expiratory pressure as an adjunct to chest physiotherapy in the treatment of cystic fibrosis. Thorax. 1986;41:951-954.

(18) Downs AM. Physiological basis for airway clearance techniques. In: Frownfelter D, Dean E, eds. Principles and Practice of Cardiopulmonary Physical Therapy. 3rd ed. St Louis, Mo: Mosby Inc; 1996:321.

(19) Milic-Emili J, Henderson JAM, Dolovich MB, et al. Regional distribution of inspired gas in the lung. J Appl Physiol. 1966;21:749-759.

(20) Fowler WS. Lung function studies, III: uneven pulmonary ventilation in normal subjects and in patients with pulmonary disease. J Appl Physiol. 1949;28:399.

(21) Van Liew HD, Mahajan RK. Display of the alveolar plateau of single-breath tests in "dilution index" format. J Appl Physiol 1989;67: 1699-1703.

(22) Gibson LE, Cooke RE. A test for concentration of electrolytes in sweat in cystic fibrosis of the pancreas utilizing pilocarpine iontophoreses. Pediatrics. 1959;23:545-549.

(23) American Thoracic Society. Standardization of spirometry: 1994 update. Am J Respir Crit Care Med. 1995;152:1107-1136.

(24) American Thoracic Society. Lung function testing: selection of reference values and interpretative strategies. Am Rev Respir Dis. 1991;144:1202-1218.

(25) Knudson RJ, Slatin RC, Lebowitz MD, Burrows B. The maximal expiratory flow-volume curve: normal standards, variability, and the effects of age. Am Rev Respir Dis. 1976; 113:587-600.

(26) Knudson RJ, Lebowitz MD, Holberg CJ, Burrows B. Changes in the normal expiratory flow-volume curve with growth and aging. Am Rev Respir Dis. 1983;127:725-734.

(27) Anthonisen NR, Danson J, Robertson PC, Ross WRD. Airway closure as a function of age. Respir Physiol. 1969;8:58-65.

(28) Olgilvie CM, Forster RE, Blackemore WS, Morton JW. A standardized breath holding technique for the clinical measurement of diffusing capacity of the lung tot carbon monoxide. J Clin Invest. 1957;36: 1-17.

(29) Buist AS, Ross BB. Quantitative analysis of the alveolar plateau in the diagnosis of early airway obstruction. Am Rev Respir Dis. 1973;108: 1078-1087.

(30) McFadden ER, Holmes B, Kiker R. Variability of closing volume measurements in normal man. Am Rev Respir Dis. 1975;111:135-140.

(31) Dubois AB, Botello SY, Bedell GN, et al. A rapid plethysmo-graphic method for measuring thoracic gas volume: a comparison with a nitrogen washout method for measuring functional residual capacity in normal subjects. J Clin Invest. 1956;35:322-326.

(32) Dubois AB, Botello SY, Comroe JH. A new method for measuring airway resistance in man using a body plethysmograph: values in normal subjects and in patients with respiratory disease. J Clin Invest. 1956;35:327-335.

(33) Crapo RO, Morris AH, Clayton PD, Nixon CR. Lung volumes in healthy nonsmoking adults. Bull Eur Physiopathol Respir. 1982;18: 419-425.

(34) McCarthy DS, Craig DB, Cherniack RM. Intraindividual variability in maximal expiratory flow-volume and closing volume in asymptomatic subjects. Am Rev Respir Dis. 1975;112:407-411.

(35) Nickerson BG, Lemen RJ, Gerdes CB, et al. With-in subject variability and percent change for significance of spirometry in normal subjects and in patients with cystic fibrosis. Am Rev Respir Dis. 1980;122: 859-866.

(36) Cooper PJ, Robertson CF, Hudson IL, Phelan PD. Variability of pulmonary function tests in cystic fibrosis. Pediatr Pulmonol. 1990;8: 16-22.

(37) Van Muylem A, Baran D. Overall and peripheral inhomogeneity of ventilation in patients with stable cystic fibrosis. Pediatr Pulmonol. 2000;30:3-9.

JC Darbee, PT, PhD, is Assistant Professor, Department of Rehabilitation Sciences, Division of Physical Therapy, College of Health Sciences, University of Kentucky, 900 S Limestone St, Lexington, KY 40536 (USA) ([e-mail]). Address all correspondence to Dr Darbee. Dr Darbee was a doctoral candidate in the Department of Physical Therapy, Exercise and Nutrition Sciences, University of Buffalo, The State University of New York, Buffalo, NY, at the time of this study.

PJ Ohtake, PT, PhD, is Associate Professor, Department of Rehabilitation Sciences, University of Buffalo, The State University of New York.

BJB Grant, MD, is Professor, Departments of Medicine, Physiology and Biophysics, and Social and Preventive Medicine, University of Buffalo, The State University of New York. Dr Grant is also Division Head of Pulmonary, Critical Care and Sleep Medicine, University of Buffalo, The State University of New York, and Veteran Affair's Medical Center, Buffalo, NY.

FJ Cerny, PhD, is Chair and Associate Professor, Department of Physical Therapy, Exercise and Nutrition Sciences, University of Buffalo, The State University of New York.

This article was received August 20, 2003, and was accepted November 24, 2003.