Facial and Ocular Deposition of Nebulized Budesonide

CHEST

Original Research RESPIRATORY ADJUNCT THERAPY

Facial and Ocular Deposition of Nebulized Budesonide* Effects of Face Mask Design Keith W. Harris, MD; and Gerald C. Smaldone, MD, PhD, FCCP

Background: In vivo case reports and in vitro studies have indicated that aerosol therapy using face masks can result in drug deposition on the face and in the eyes, and that face mask design may affect drug delivery. Objective: To test different mask/nebulizer combinations for budesonide, a nebulized steroid used to treat pediatric patients with asthma. Methods: Using high-performance liquid chromatography, drug delivery (inhaled mass), facial, and ocular deposition of budesonide aerosols were studied in vitro using a ventilated face facsimile (tidal volume, 50 mL; rate, 25 breaths/min, duty cycle 0.4), a tight-fitting test mask, a standard commercial mask, and a prototype mask designed to optimize delivery by reducing particle inertia. Nebulizer insertion into the mask (front loaded vs bottom loaded) was also tested. Particle size was measured by cascade impaction. Pari LC Plus (PARI Respiratory Equipment; Midlothian, VA) and MistyNeb (Allegiance; McGaw Park, IL) nebulizers were tested. Results: Inhaled mass for tight-fitting and prototype masks was similar (13.2 ⴞ 1.85% vs 14.4 ⴞ 0.67% [percentage of nebulizer charge], p ⴝ 0.58) and significantly greater than for the commercial mask (3.03 ⴞ 0.26%, p ⴝ 0.005). Mask insertion of nebulizer was a key factor (inhaled mass: front loaded vs bottom loaded, 8.23 ⴞ 0.18% vs 3.03 ⴞ 0.26%; p ⴝ 0.005). Ocular deposition varied by an order of magnitude and was a strong function of mask design (4.77 ⴞ 0.24% vs 0.35 ⴞ 0.05%, p ⴝ 0.002, tight fitting vs prototype). Particle sizes (7.3 to 9 ␮m) were larger than previously reported for budesonide. Conclusions: For pediatric breathing patterns, mask design is a key factor defining budesonide delivery to the lungs, face, and eyes. Front-loaded nebulizer mask combinations are more efficient than bottom-loaded systems. (CHEST 2008; 133:482– 488) Key words: deposition; high-performance liquid chromatography; nebulized budesonide Abbreviation: MMAD ⫽ mass median aerodynamic diameter

the advent of aerosolized therapy, anecdotal S ince reports have demonstrated that aerosols can deposit in the eyes of adult1–3 and pediatric1,4 – 8 patients, causing clinically relevant pupillary dilatation and glaucoma. The face mask, traditionally viewed as a simple interface between the aerosol device and the patient, is now recognized as a key factor affecting both aerosol delivery to the lungs,9 –12 as well as deposition of particles on the face and in the eyes.5,11–14 In addition to the acute effects of bronchodilators, the potential long-term effects of aerosolized steroids with ocular deposition combined with the need to ensure optimal delivery to the lungs 482

for efficacy suggest that face mask design may also be important for these agents. A study13 using models has noted that although a tight-fitting mask improves efficiency of delivery, ocular deposition is enhanced. A study14 of face mask design has indicated that inertial effects along the edge of the mask are important in regulating facial deposition particularly in the eyes. Budesonide (Pulmicort respules; AstraZeneca LP; Westborough, MA), is a suspension and hence may behave differently than a solution of albuterol or ipratropium. The purpose of the present in vitro study is to test the influence of face mask design on the behavior of nebulized budesonide. Original Research

Figure 1. The face model.

Materials and Methods Face Model To assess the factors defining drug delivery via face mask, it is necessary to have a face/face mask model that ventilates like a patient. A custom-made model of a 2-year-old child’s face (Consulting Group; Cambridge Technology Centre; Melbourn, UK) was used to mimic the child (Fig 1). The dimensions of the face were 22 cm high, 13 cm wide, and 8 cm deep, with an 18 mm in diameter orifice as a substitute for a mouth. Aerosol was generated from various nebulizers with a Pro-Neb Ultra compressor (PARI Respiratory Equipment; Midlothian, VA). The aerosol was inhaled using a piston pump (Harvard Apparatus; South Natick, MA; Fig 2) with a pediatric breathing pattern (tidal volume, 50 mL; frequency, 25 breaths/min; and duty cycle of 0.4). Delivered dose to the patient (inhaled mass) was determined by budesonide captured on the inhaled mass filter (Pari; Starnberg, Germany) placed between the mannequin and the piston pump. Aerosol deposition over the eyes was determined by extracting drug from filters placed over the eye regions (Fig 3) as well as the rest of the face and the face mask (see below). *From the Division of Pulmonary, Critical Care and Sleep Medicine, Department of Medicine, State University of New York, Stony Brook, NY. Dr. Harris has no conflict of interest to disclose. Dr. Smaldone is a paid consultant for PARI Respiratory Equipment, Inc, has consulted to AstraZeneca, and has licensed patents to PARI on face mask design. The development of this technology was supported in part by AstraZeneca and PARI Respiratory Equipment, Inc. Manuscript received July 24, 2007; revision accepted October 29, 2007. Reproduction of this article is prohibited without written permission from the American College of Chest Physicians (www.chestjournal. org/misc/reprints.shtml). Correspondence to: Gerald C. Smaldone, MD, PhD, Division of Pulmonary, Critical Care and Sleep Medicine, Department of Medicine, State University of New York, Stony Brook, NY 11794-8172; e-mail: [email protected] DOI: 10.1378/chest.07-1827 www.chestjournal.org

Figure 2. Sketch of experimental configuration. Ventilation of the face is provided by the piston pump, drug particles passing through the mouth are captured on the inhaled mass filter representing aerosol directed to the respiratory tract; upper panel shows front-loaded configuration; lower panel shows bottomloaded configuration.

Drug Delivery, Mass Balance, and Analysis Budesonide at a dose of 0.5 mg in 2.0 mL (0.25 mg/mL) was used. The amount of drug placed in the nebulizer at the beginning of each run is referred to as the nebulizer charge. For each experiment, the nebulizer was charged and run until dry three times. With each nebulizer run, the inhaled mass filter was changed while the facial and mask deposition was allowed to accumulate on the mannequin and face mask. Then, budesonide was extracted from each inhaled mass filter (and filter holder) in a beaker containing 25 mL of ethanol, with fluocinolone (40 mg/L) serving as an internal standard. The final inhaled mass value represents the sum of all three inhaled mass filters. Aerosol deposition over the eyes was determined after the three nebulizer runs by extracting drug from filters placed over the eye regions (Fig 3). Drug activity over the rest of the face (ie, excluding the eye region) and over the face mask (inner surface) was measured after completion of the three runs by washing each with 25-mL aliquots of ethanol containing fluocinolone and analyzing all solutions for budesonide by high-performance liquid chromatography according to AstraZenca method D-390 – 01.15 Budesonide delivery to the “patient” was reported as percentage of the cumulative nebulizer charge (1,500 ␮g), which was distributed in the eye filters, on the face, and the three inhaled mass filters. The measurements from the summed inhaled mass, cumulative face, and mask deposition represent a single data point. This entire process was performed in duplicate for each experimental configuration. CHEST / 133 / 2 / FEBRUARY, 2008

483

Figure 3. Filters over the eye regions, which allowed quantification of deposited aerosol on the face and in the “eyes.”

Face Mask and Nebulizer Configurations Face mask and nebulizer configurations are listed in Table 1. Masks are pictured in Figure 4. The Laerdal mask (Laerdal Medical Corporation; Wappingers Falls, NY) represented a paradigm of a tight-fitting mask designed to minimize facemask leaks. A Pari LC Plus jet nebulizer was attached straight in or “front loaded” with particles from the nebulizer directed toward the face (Fig 2, upper panel). In a previous study,13 this configuration delivered aerosol with the highest efficiency to the patient but with tight-fitting masks resulted in significant deposition in the eyes. A typical, commercially available “bottom-loaded” nebulizer configuration is depicted in Figure 2 (lower panel). This system combined a Salter mask (Salter Labs; Arvin, CA), displayed in Figure 4, and a MistyNeb nebulizer (Allegiance; McGaw Park, IL). In this configuration, the aerosol from the nebulizer changed direction (90°) before moving toward the face. The third configuration consisted of a front-loaded commercial prototype mask (PARI Respiratory Equipment) designed to incorporate the attributes of a tight-fitting front-loaded system

Table 1—Mask/Nebulizer Combinations Mask

Nebulizer*

Laerdal Salter Prototype Prototype

LC Plus MistyNeb LC Plus MistyNeb

*Pari Pro-Neb Ultra Compressor (PARI Respiratory Equipment) used with all nebulizers. 484

Figure 4. The different face masks: tight-fitting (Laerdal Medical Corporation), upper panel; typical commercial mask (Salter Labs), middle panel; and commercial prototype mask, a modified version of the Pari Bubbles the Fish mask (PARI Respiratory Equipment), lower panel.

(highest efficiency) with modifications designed to minimize facial and eye deposition.14 The prototype was placed on the face but not “sealed” like the Laerdal resuscitation mask. Cutouts along the edge of the mask in the region of the eyes controlled the velocity of linear jets at leaks near the eyes. Two nebulizers were tested with this mask: Pari LC Plus and MistyNeb. Each nebulizer was placed in a front-loaded position to maximize aerosol delivery similar to the Laerdal configuration. A commercially available elbow (PARI Respiratory Equipment) was utilized between the mask and MistyNeb to connect that nebulizer to the mask. Aerodynamic Particle Distribution The distribution of nebulized particles for the different aerosols was determined using a seven-stage, low-flow (2 L/min) cascade impactor (Marple; Thermo Electron Corporation; Waltham, MA) expressed as the mass median aerodynamic diameter (MMAD) of budesonide produced by each configuration.13–15 The cascade impactor was placed between the nebulizer and the piston pump, and particles were sampled as they exited the nebulizer (Fig 5, upper panel for front-loaded nebulizer, lower panel for bottomloaded nebulizer). Budesonide captured on the filters for each stage was extracted and analyzed by high-performance liquid chromatography as described above. In summary, therefore, these experiments were designed to define the following: 1. What is the maximal delivery of budesonide to a young child using a tight-fitting mask? What is the potential for Original Research

Figure 5. Cascade impaction setup: front-loaded system (upper panel); bottom-loaded system (lower panel).

facial and eye deposition of this configuration? Can a prototype design match the efficiency of the tight-fitting system and minimize facial and eye deposition? 2. What is the typical delivery of budesonide to a young child using a typical commercially available mask and nebulizer? What is the potential for facial and eye deposition of this configuration? How does it compare to the prototype? 3. How does efficiency of a bottom-loaded system compare to a front-loaded system? 4. What is the role of the nebulizer? The experimental configurations listed in Table 1 were designed to answer these questions. Based on previous studies using solutions as the drug formulation,13,14 the Laerdal/Pari LC Plus nebulizer represents the most efficient, tight-fitting system. Commercial systems are represented by the Salter/MistyNeb combination. The prototype/Pari LC Plus combination assessed factors defining facial and eye deposition. Finally, the nebulizer effect was determined by testing both nebulizers with the prototype mask. www.chestjournal.org

Statistics Mass balance calculations were performed to quantify deposition on the face, eye region, the face mask, and inhaled mass filter, and results are reported as percentage of nebulizer charge. Descriptions of the data are presented in terms of means and SEs. The data were normalized by square root transformations for the two-group comparisons, but the data are described in their original units, percentages. The t test was used for the normalized data for the comparisons, and the resulting p values are reported. Statistical software was used to analyze the data (SPSS version 15; Chicago, IL; and StatXact; Cytel Statistical Software and Services; Cambridge, MA).

Results Inhaled Mass, Facial, and Eye Deposition Drug delivery (inhaled mass) for the tight-fitting (Laerdal) and prototype mask with the Pari LC CHEST / 133 / 2 / FEBRUARY, 2008

485

Table 2—Effect of Mask Design on Inhaled Mass, Facial, and Eye Deposition* Variables

Laerdal/LC Plus

Prototype/ LC Plus

p Value

Inhaled mass Facial deposition† Eye deposition Mask deposition

14.4 ⫾ 0.67 6.51 ⫾ 0.13 4.77 ⫾ 0.24 1.06 ⫾ 0.07

13.2 ⫾ 1.85 1.33 ⫾ 0.24 0.35 ⫾ 0.05 1.99 ⫾ 0.16

0.58 0.006 0.002 0.028

*Data are presented as mean ⫾ SE. †Entire face including eyes.

Plus nebulizer are listed in Table 2. Inhaled mass for both systems was similar (Laerdal vs Prototype, 14.4 ⫾ 0.67% and 13.2 ⫾ 1.85%, respectively; p ⫽ 0.58). Facial deposition was markedly higher for the Laerdal mask at 6.51 ⫾ 0.13% compared to the prototype 1.33 ⫾ 0.24% (p ⫽ 0.006). For the Laerdal mask, facial deposition was mainly focused around the eyes, 4.77 ⫾ 0.24%, an order of magnitude greater than that measured for the prototype, 0.35 ⫾ 0.05% (p ⫽ 0.002). Deposition on the mask itself was less for the Laerdal, 1.06 ⫾ 0.07% compared to the prototype, 1.99 ⫾ 0.16% (p ⫽ 0.028).

Table 4 —MMAD for Budesonide Aerosols Using Different Mask/Nebulizer Combinations Mask Type

Nebulizer

Laerdal Prototype

LC Plus

MMAD, ␮m 7.8 7.6 9.0 7.3

MistyNeb Salter

onide suspension, MMAD ranged from 7.3 to 9.0 ␮m. Similar aerosols were seen for the front-loaded Pari LC Plus using either Laerdal or prototype masks. The largest particles (9.0 ␮m) were observed using the MistyNeb device in the front-loaded position, with a decrease in MMAD to 7.3 ␮m for the same nebulizer in the bottom-loaded position. Role of the Nebulizer As shown in Table 5, using the Prototype mask and nebulizers of different design did not significantly affect drug delivery (inhaled mass), deposition on the face, eyes, or the mask. When combined with the data in Table 3, mask design appears to be the dominant factor affecting aerosol delivery, and facial and eye deposition.

Bottom-Loaded vs Front-Loaded Nebulizers Using the MistyNeb nebulizer, the front-loaded position (prototype mask) was significantly more efficient than the bottom-loaded configuration (Salter mask) as shown in Table 3 (8.23 ⫾ 0.18% and 3.03 ⫾ 0.26%, p ⫽ 0.005). In spite of the large differences in inhaled mass, facial and eye deposition for the prototype mask design were reduced compared to the Salter commercial configuration (2.1 ⫾ 0.06% vs 3.07 ⫾ 0.09%, p ⫽ 0.011, and 0.53 ⫾ 0.03% vs 1.35 ⫾ 0.04%, p ⫽ 0.004, respectively). However, mask deposition was more variable between the two systems (2.21 ⫾ 0.33% vs 6.52 ⫾ 1.18%, p ⫽ 0.054).

Discussion

MMADs were measured for all mask/nebulizer configurations and are listed in Table 4. For budes-

The present in vitro study demonstrates that for nebulization of budesonide suspension, face mask design is a key factor in drug delivery to the patient and to the face, particularly in the region of the eyes. While our study used an in vitro model, it is clear from human case reports that ocular deposition can occur. Using models, we are attempting to define the factors that govern drug delivery to the lungs and the face under clinically relevant circumstances. We chose the 50-mL tidal volume because we believed that it created a worst-case scenario with most of the delivered gases vented through and around the face mask so an optimized design should handle 50 mL as a first step. Using the same model and pattern of breathing, Sangwan et al13 demonstrated that in-

Table 3—Effect of Nebulizer Orientation on Drug Delivery*

Table 5—Effect of Nebulizer on Drug Delivery, Both Devices in Front-Loaded Configuration*

Particle Size and Drug Delivery

Variables Inhaled mass Facial deposition† Eye deposition Mask deposition

Salter/MistyNeb Prototype/MistyNeb (Bottom Loaded) (Front Loaded) p Value 3.03 ⫾ 0.26 3.07 ⫾ 0.09 1.35 ⫾ 0.04 6.52 ⫾ 1.18

*Data are presented as mean ⫾ SE. †Entire face including eyes. 486

8.23 ⫾ 0.18 2.1 ⫾ 0.06 0.53 ⫾ 0.03 2.21 ⫾ 0.33

0.005 0.011 0.004 0.054

Variables

Prototype/LC

Prototype/ MistyNeb

p Value

Inhaled mass Facial deposition† Eye deposition Mask deposition

13.21 ⫾ 1.85 1.33 ⫾ 0.24 0.35 ⫾ 0.05 1.99 ⫾ 0.16

8.23 ⫾ 0.18 2.1 ⫾ 0.06 0.53 ⫾ 0.03 2.21 ⫾ 0.33

0.099 0.102 0.099 0.610

*Data are presented as mean ⫾ SE. †Entire face including eyes. Original Research

haled mass for saline solution test aerosols ranged from 2 to 6% of the nebulizer charge. In the present study, budesonide delivery varied from 3 to 14%, indicating that with this formulation aerosol delivery can be enhanced compared to solutions. The present study, as well as previous studies13,14 using solutions, demonstrates that the major factor affecting aerosol delivery is the front-loaded nebulizer position. We did not detect significant differences between nebulizers in this study. It is possible that significant differences would have been detected if we had performed more experiments, but for this breathing pattern the major factor influencing delivery was the front-loaded position. Both the Pari and MistyNeb, nebulizers, when front loaded, delivered between 8 to 14% to the inhaled mass filter (Table 2), and interfacing the MistyNeb to a bottom-loaded Salter mask reduced inhaled mass to 3% (Table 3). While a front-loaded delivery system was more efficient than a bottom-loaded system, facial and eye deposition were significant. For example, in Table 2 ocular deposition was 30% of drug delivery to the patient. A previous study14 has suggested increased particle inertia along the edge of the mask as an explanation for excess facial deposition. The “tight-fitting mask” does not provide a perfect seal but facilitates small leaks with high local velocities that enhance facial and ocular deposition. Therefore, an improved delivery configuration should include a front-loaded system with a mask that reduces particle inertia in the region of the eyes. The prototype mask is a loose-fitting mask that incorporates several modifications designed to reduce the ballistic properties of the particles. The most important are the large cutouts around the edge near the eyes.14 The prototype mask markedly reduced facial and eye deposition while maintaining high inhaled mass (Table 2). When compared to the bottom-loaded Salter system and using the same nebulizer (MistyNeb), the prototype front-loaded system more than doubled inhaled mass with significantly less facial and eye deposition (Table 3). As shown in Table 4, particle distributions for each configuration were similar, indicating that for aerosolization of budesonide the efficiency of a given system can be improved with the prototype mask and that this improvement is not attributable to changes in particle size. Studying the face mask as a device rather than as a simple connector is new and evolving. As we have found with nebulizers, drug delivery is a function of the device and the formulation and, for face masks, the interaction between the mask and the face. For budesonide, some of the results were unexpected, particularly the magnitude of facial deposition and the inhaled mass, and the size of the aerosol particles. Particle inertia can be affected by several parameters www.chestjournal.org

common to the delivery of aerosols using face masks: particle size and the flow of gas within the mask. For example, we found that increased compressor flow can be a factor in facial deposition and that the modifications made in our prototype design minimized these effects.14 In addition, we have found that during aerosol delivery of solutions with face masks, MMADs are increased when compared to aerosol delivery with a mouthpiece. For example, using the Pari LC Plus nebulizer, Diot at al16 measured MMAD of 3 ␮m for Pulmozyme aerosols inhaled via a mouthpiece. For face mask aerosol delivery, using the same nebulizer, we have reported particle sizes with MMADs ranging from 5.8 to 6.6 ␮m14; and, in the present study using Pulmicort suspension, we found that the Pari LC Plus delivered particles of 7.7 ␮m. Similar comparisons can be made for the MistyNeb nebulizer. These data indicate that there are two factors affecting MMADs reported in these studies: the use of face mask vs mouthpiece and the formulation (solution vs suspension). For tight-fitting face masks, facial deposition for aerosols from solutions averaged 2%, inhaled mass 6%13; for budesonide aerosols, with increased MMAD, 6.5% and 14.4%, respectively. All other things being equal, facial deposition will be proportional to particle size and mask design can minimize facial and ocular deposition. Our group tested nebulizer mouthpiece combinations as part of the preclinical assessment of delivery systems prior to the clinical trials of budesonide in the United States.15 Using a similar in vitro setup, but without the face mask, the Pari LC Plus nebulizer (the device used in the Pulmicort clinical trials) averaged 18% inhaled mass. In the present article, the same nebulizer using a front-loaded mask delivered 14%. These results suggest that a child using a mask will receive a typical dose to that of a nebulizer using a mouthpiece. However, switching to a bottomloaded configuration could reduce delivery to the range of 3%. Our face model cannot be expected to perfectly reproduce all possible “faces” or, in fact, any living face. It was designed to allow the measurement of worse-case scenarios and, with mask design, to minimize them. We specifically chose the Laerdal mask for comparison to the prototype for two reasons: a tight-fitting system is often recommended for use with budesonide suspension (original package insert), and it provides high inhaled mass. The Salter mask was chosen to provide a bottom-loaded system (the most common clinical configuration) for comparison with the front-loaded prototype. The respective nebulizers were used to evaluate efficiency (breath-enhanced [Pari], constant flow [MistyNeb]), CHEST / 133 / 2 / FEBRUARY, 2008

487

compatibility with masks, and because they represent the spectrum of typical commercial nebulizers.15 Are these bench observations important to clinical practice? Aerosols formulated as solutions or suspensions can be delivered using mouthpiece or mask, and both forms of delivery have demonstrated efficacy in clinical trials, particularly for Pulmicort respules using the Pari LC Plus nebulizer. Deposition studies in children using this formulation are not available; however, MMADs for both mouth breathing and face mask breathing indicate that a significant portion of the aerosol may deposit in the upper airways and possibly on the face. In vivo data, using solutions and mouthpiece, have indicated that MMADs of 3 ␮m in children result in 50% upper airway deposition.16 We expect, therefore, that budesonide aerosols with greater MMAD will have significant upper airway deposition. Are facial and ocular deposition clinically relevant considerations? No studies measuring in vivo deposition on the face are available using budesonide suspension. However, Erzinger and colleagues,5 in a unique in vivo study, measured significant deposition on the face in children inhaling salbutamol aerosols using tight- and loose-fitting face masks. They found that facial deposition was between 2.6% and 8.4%, findings similar to our in vitro data for tight-fitting face masks using both saline solution13,14 and budesonide suspension. These effects are minimized in vitro by modifying face mask design. Our data and the data of Erzinger et al5 begin to quantify the effects seen clinically in the anecdotal reports of facial and eye deposition during aerosol therapy. The study of Erzinger et al5 also indicates that mask design is important for aerosol systems that do not use compressed gas. Those authors found facial deposition of similar magnitude to tight-fitting masks and nebulizers in children treated with a metereddose inhaler (Ventolin; Allen Hanburys; Sydney, NSW, Australia) using a valved holding chamber (Aerochamber; Trudell Medical; London, ON, Canada). Their study suggests that aerosol particles may gain inertia during expiration because the holding chamber is likely not pressurized after the initial puff from the inhaler during inspiration.5 The findings of our study and that of Erzinger et al5 suggest that clinical studies of aerosolized drugs should consider the face mask an integral factor in aerosol delivery along with the nebulizer and the formulation. Principles of face mask design are likely important in aerosol delivery to adults as well as children and during other forms of aerosol therapy involving face masks, such as noninvasive ventilation.17 Side effects associated with long-term aerosol use are largely undefined, but the reported cases of patients suffering from acute pupillary dilatation secondary to bronchodilator 488

deposition in the eyes indicate the potential for longterm effects from all aerosolized drugs. We have found that mask design can minimize this potential by minimizing facial deposition while ensuring efficient drug delivery to the patient. In conclusion, the choice of face mask can be an important factor in determining the delivery of nebulized budesonide in children. Appropriate combinations of face mask and nebulizer are needed to maximize inhaled mass and minimize facial deposition. ACKNOWLEDGMENT: The authors thank Roger Grimson, PhD, for statistical analysis, and Lorraine Morra, BS, and Akbar Shah, MD, for technical assistance.

References 1 Bisquerra RA, Botz GH, Nates JL. Ipratropium-bromideinduced acute anisocoria in the intensive care setting due to ill-fitting face masks. Respir Care 2005; 50:1662–1664 2 Iosson N. Images in clinical medicine: nebulizer-associated anisocoria. N Engl J Med 2006; 354:e8 3 Hall SK. Acute angle-closure glaucoma as a complication of combined beta-agonist and ipratropium bromide therapy in the emergency department. Ann Emerg Med 1994; 23:884 – 887 4 Brodie T, Adalat S. Unilateral fixed dilated pupil in a well child [letter]. Arch Dis Child 2006; 91:961 5 Erzinger S, Schueepp KG, Brooks-Wildhaber J, et al. Facemasks and aerosol delivery in vivo. J Aerosol Med 2007; 20(suppl 1):S78 –S84 6 Ryan CA. Styrofoam cup, ipratropium bromide, and anisocoria [letter]. Clin Pediatr (Phila) 1999; 38:318 7 Woelfle J, Zielen S, Lentze MJ. Unilateral fixed dilated pupil in an infant after inhalation of nebulized ipratropium bromide. J Pediatr 2000; 136:423– 424 8 Mulpeter KM, Walsh JB, O’Connor M, et al. Ocular hazards of nebulized bronchodilators. Postgrad Med J 1992; 68:132–133 9 Amirav I, Newhouse MT. Aerosol therapy with valved holding chambers in young children: importance of the facemask seal. Pediatrics 2001; 108:389 –394 10 Morton RW, Mitchell JP. Design of facemasks for delivery of aerosol-based medication via pressurized metered dose inhaler with valved holding chamber: key issues that affect performance. J Aerosol Med 2007; 20(suppl 1):S29 –S45 11 Nikander K, Berg E, Smaldone GC. Jet nebulizers versus pressurized metered dose inhalers with valved holding chambers: effects of the facemask on aerosol delivery. J Aerosol Med 2007; 20(suppl 1):S46 –S58 12 Smaldone GC, Berg E, Nikander K. Variation in pediatric aerosol delivery: importance of facemask. J Aerosol Med 2005; 18:354 –363 13 Sangwan S, Gurses BK, Smaldone GC. Facemasks and facial deposition of aerosols. Pediatr Pulmonol 2004; 37:447– 452 14 Smaldone GC, Sangwan S, Shah A. Facemask design, facial deposition, and delivered dose of nebulized aerosols. J Aerosol Med 2007; 20(suppl 1):S66 –S77 15 Smaldone GC, Cruz-Rivera M, Nikander K. In vitro determination of inhaled mass and particle distribution for budesonide nebulizing suspension. J Aerosol Med 1998; 11:113–125 16 Diot P, Palmer LB, Smaldone A, et al. RhDNase I aerosol deposition and related factors in cystic fibrosis. Am J Respir Crit Care Med 1997; 156:1662–1668 17 Hess DR. The mask for noninvasive ventilation: principles of design and effects on aerosol delivery. J Aerosol Med 2007; 20(suppl 1):S85–99 Original Research