Published ahead of print on October 29, 2004, doi:10.1164/rccm.200407-895OC
© 2005 American Thoracic Society doi: 10.1164/rccm.200407-895OC
Multiple-Breath Washout as a Marker of Lung Disease in Preschool Children with Cystic FibrosisPortex Respiratory Unit, Institute of Child Health; Cardiorespiratory and Critical Care Division, Great Ormond Street Hospital for Children; Department of Paediatric Respiratory Medicine, Royal Brompton Hospital; Department of Paediatrics, Kings' College Hospital; Department of Paediatrics, Lewisham University Hospital; and Department of Paediatrics, Royal London Hospital, London, United Kingdom; and Queen Silvia Children's Hospital, Sahlgrenska University Hospital, Göteborg, Sweden Correspondence and requests for reprints should be addressed to Paul Aurora, Portex Respiratory Unit, Institute of Child Health, 30 Guilford Street, London WC1N 1EH, UK. E-mail: p.aurora{at}ich.ucl.ac.uk
Sensitive measures of lung function applicable to young subjects are needed to detect early cystic fibrosis (CF) lung disease. Forty children with CF aged 2 to 5 years and 37 age-matched healthy control subjects performed multiple-breath inert gas washout, plethysmography, and spirometry. Thirty children in each group successfully completed all measures, with success on first visit being between 68 and 86% for all three measures. Children with CF had significantly higher lung clearance index (mean [95% CI] difference for CF control 2.7 [1.9, 3.6], p < 0.001) and specific airway resistance (1.65 z-scores [0.96, 2.33], p < 0.001), and significantly lower forced expired volume in 0.5 seconds (–0.49 z-scores [–0.95, –0.03], p < 0.05). Abnormal lung function results were identified in 22 (73%) of 30 children with CF by multiple-breath washout, compared with 14 (47%) of 30 by plethysmography, and 4 (13%) of 30 by spirometry. Children with CF who were infected with Pseudomonas aeruginosa had significantly higher lung clearance index, but no significant difference in other lung function measures, when compared with noninfected children. Most preschool children can perform multiple-breath washout, plethysmography, and spirometry at first attempt. Multiple-breath washout detects abnormal lung function in children with CF more readily than plethysmography or spirometry.
Key Words: child • cystic fibrosis • lung function • preschool There is increasing emphasis on the detection and aggressive treatment of early cystic fibrosis (CF) lung disease, resulting in increased interest in lung function testing of infants and preschool children with CF. Techniques for measurement of lung function in infants are long established, and have been recently standardized (1), unlike lung function testing in the preschool age group. There are two important issues in lung function testing of preschool children with CF. First, most lung function tests performed in older children require a degree of active cooperation, which is beyond many preschool children. Second, parameters derived from maximal expiratory maneuvers (spirometry), which have traditionally been used to monitor lung function in patients with CF, may be of less value in younger patients. Although the FEV1 is still recognized to be a good predictor of prognosis in subjects with moderate to severe CF lung disease (2–4), many younger school-age subjects have FEV1 results in the normal range (5). It is now clear that many preschool children are able to perform spirometry, although modifications to both measurement techniques and outcome measures are necessary (6). In addition, two alternative techniques for studying this population have recently been described. One is multiple-breath inert gas washout (MBW), a method for estimating the efficiency of gas mixing in the lung. A recent report suggests that indices derived from sulfur hexafluoride (SF6) MBW may be more sensitive than spirometry in detecting airway disease in older children with CF (7). The second is the measurement of specific airway resistance (sRaw) by body plethysmography (8). The aim of this study was to compare MBW, body plethysmography, and spirometry in preschool children with CF and healthy preschool children. The selected outcome measures were the lung clearance index (LCI) from MBW, sRaw from body plethysmography, and the forced expiratory volume in 0.5 seconds (FEV0.5) and mean expiratory flow between 25% and 75% of forced vital capacity (FEF25–75) from spirometry. FRC was also reported from MBW, and FVC from spirometry is reported in the online supplement. The study population was a cohort of 2- to 5-year-old children with CF, and an age-matched group of healthy children. Some of the results of this study have been previously reported in abstract form (9).
A full description of study methods is provided in the online supplement. Subjects with CF, diagnosed on standard criteria (10), were primarily recruited from the CF clinic at Great Ormond Street Hospital, with a minority recruited from the other four centers participating in the London Cystic Fibrosis Collaboration (11, 12). Healthy control children were primarily recruited from local schools and playgroups, although a minority were friends or siblings of children with CF. Children who had been diagnosed with asthma, who were taking antiasthma medication, who had a past history of pneumonia, pertussis, or tuberculosis infection, who were born before 34 weeks' gestation, who had been hospitalized for respiratory infection, who had neuromuscular or bone disease likely to affect respiration, or who had congenital cardiac disease requiring treatment were excluded from this study. All subjects were aged between 2 and 6 years at time of testing. For children with CF, information on current respiratory infection was obtained from the clinical center. Current bacterial infection was defined as at least one positive culture of an organism from the three most recent samples taken before testing. The study was approved by the North Thames Region Multicentre Research Ethics Committee, and by the Research Ethics Committees of the collaborating centers. Written informed consent was obtained from the parents of all subjects. All tests were performed at the preschool lung function laboratory at the Institute of Child Health, London, between January 2001 and November 2003. Children were measured when well: i.e., children with CF did not have a pulmonary exacerbation, and control children had no evidence of acute respiratory infection. Children with CF received their usual airway clearance therapy on the morning before testing, and lung function testing commenced at some time between 9:00 A.M. and 2:00 P.M. All children performed MBW, plethysmography, and spirometry, in that order, following a set protocol. Children who were unable successfully to complete this protocol were asked to return for a repeat visit approximately 6 months later. The technique for performing MBW in school-age children has been described previously (7). In the current study, children were required to breathe by a Rendell-Baker size 2 facemask (Ambu Intl., Bath, Avon, UK), which was applied to the child's face using therapeutic putty to form an airtight seal. Flow was measured by a Fleisch No.1 pneumotachometer, and gas concentrations were measured by a respiratory mass spectrometer (AMIS 2000; Innovision A/S, Odense, Denmark). During the test the room was darkened and the child was encouraged to watch a children's entertainment video with the aim of distracting the subject and encouraging regular tidal breathing. Each test consisted of two phases. During the wash-in phase the subject inspired a dry air mixture containing 4% SF6, 4% helium, 21% oxygen, and balance nitrogen. The SF6 was the marker gas used for calculation of gas mixing indices reported in this study. Wash-in was continued until the inspiratory and expiratory SF6 concentrations were stable and equal to within 0.1%, plus another 10 seconds. At this moment the bias flow was stopped during expiration and washout was started using room air. The washout phase continued until the end-tidal SF6 concentration was below 0.1%. The LCI was calculated by dividing the cumulative expired volume by the FRC, as described previously (13, 14). The mean LCI and mean FRC from the first three technically acceptable washouts were calculated, and are presented here. If less than three acceptable washouts were obtained, the results were not reported. The sRaw was measured with a constant volume body plethysmograph (MasterScreen Body; VIASYS Healthcare, Höchberg, Germany). The child sat alone in the plethysmograph, and was asked to wear a noseclip. The child was then asked to breathe through the mouthpiece, at a rate of at least 30 breaths per minute, while between 5 and 25 loops was recorded. If fewer than five technically acceptable loops were obtained the results were not reported. Total specific airway resistance was reported (15). Spirometry was performed using a Jaeger MasterScope spirometer (VIASYS Healthcare, Höchberg, Germany), as previously described (6).
Data Analysis Z-scores for height and weight were calculated from published reference equations (16) for British children. Z-scores for FEV0.5, FEF25–75, sRaw, and FRC were calculated from the data obtained in the healthy children reported here. For FEV0.5 and FEF25–75 a z-score of less than –1.96 was classified as abnormal. For FRC and sRaw a z-score of greater than 1.96 was classified as abnormal. The mean and SD for LCI were also calculated from the control subject population. An LCI greater than 1.96 SD above the mean was classified as abnormal. A secondary analysis, based on z-scores calculated from published reference data (15, 17, 18), is presented in the online data supplement. Within-test repeatability for LCI and for sRaw was determined by calculating the coefficient of variation (CVLCI; CVsRaw) as 100 · mean · SD-1. Results for sRaw, FEV0.5, FEF25–75, LCI, FRC, and CVLCI were compared for the CF and control groups. In a subanalysis, lung function results were compared for children with CF who were currently infected with Pseudomonas aeruginosa against those who were not. A second subanalysis, comparing lung function results with chest radiograph scores, is presented in the online supplement. Summary statistics are presented as mean and SD. Proportions were compared by chi-square test, or by Fisher exact test as appropriate. The t test was used for comparison of two groups; analysis of variance and post hoc analysis by Tukey's honestly significant difference was used for comparison of three or more groups. For all analyses a p value of 0.05 or less was regarded as significant, and the 95% confidence intervals for difference of the mean are presented where available.
Forty children with CF, with a mean (SD) age of 4.1 (0.9) years, and 37 healthy children with a mean age of 4.2 years (0.9 years), were recruited to this study. Success rates in completing MBW, plethysmography, and spirometry on the first visit are presented in Tables 1 and 2. Healthy children were significantly more likely to complete plethysmography than children with CF (p = 0.03). There was no other significant difference in success by diagnosis. Children aged less than 3 years were less likely successfully to complete plethysmography (p < 0.001) or spirometry (p < 0.001) than older children. There was no significant difference in success by age for MBW. Fifty of the 77 children were able to complete the study protocol on the first test visit.
On repeat visits a further three healthy controls and seven children with CF were able to complete the protocol, giving a total study population of 60 children (30 with CF). Twenty-two of the children with CF were homozygous for the  F508 mutation; the remainder had one F508 mutation and one other. Of these eight children, three had unknown second mutations, and one each had 1717–1G-A, 1898+1G-A, N1303K, I507, and G542X mutations. Some of the spirometry data from 44 of these 60 children have been presented previously, as part of a spirometry quality control study (6). Summary data for the final study population are presented in Table 3. Although children with CF were slightly lighter and shorter than the healthy children, these differences were not statistically significant. Mean duration of each MBW run was 59 seconds for children with CF and 62 seconds for healthy children. This difference was not significant. Children with CF tended to have a higher mean respiratory rate during the MBW than healthy children (27.5 versus 24.4 breaths · min-1, mean difference 3.1 [0, 6.2], p = 0.051). There was no difference in mean tidal volume during MBW between the two groups.
LCI, sRaw, FEV0.5, FEF25–75, and FRC were regressed against height for the healthy children. There was no relationship between LCI and height. The results of the other regression analyses are presented in the on-line supplement. Age relationships for LCI, sRaw z-score, FEV0.5 z-score, and FEF25–75 z-score are presented in Figure 1. There was no relationship between LCI and age, either in healthy children or in the children with CF. The group mean for LCI for the healthy children was 6.89 (0.44) with 95% limits of normality for LCI calculated as 6.01; 7.77. These limits are displayed in Figure 1A. Lung function results were compared for CF and control groups (Table 4). Children with CF had significantly higher LCI and sRaw and significantly lower FEV0.5 than the control children. There was no significant difference in FEF25–75 or FRC z-scores between the groups.
CVLCI was 7.8% (5.4%) in children with CF, and 5.2% (2.3%) in healthy children (mean [95% CI] difference 2.6%, [0.4%, 4.8%], p = 0.04). There was no relationship between CVLCI and age in either subject group. CVsRaw was almost identical for CF and control groups (10.5% [4.8%] versus 11% [4.9%]) with no age relationships seen. Twenty-two of the 30 children with CF had abnormal LCI. Fourteen had an abnormal sRaw, of whom 13 also had an abnormal LCI (Figure 2A). Two children with CF had an abnormal FEV0.5 (Figure 2B), and four had an abnormal FEF25–75 (Figure 2C), all of whom also had an abnormal LCI. None of the children with CF had abnormally raised FRC. In children with CF, weak but statistically significant correlations between lung function parameters were noted. LCI was positively correlated with sRaw (r2 = 0.14, p = 0.04), and negatively correlated with FEV0.5 (r2 = 0.21, p = 0.01) and FEF25–75 (r2 = 0.28, p = 0.003). sRaw was negatively correlated with FEV0.5 ([Figure 2D], r2 = 0.22, p = 0.01) and with FEF25–75 ([Figure 2E], r2 = 0.33, p = 0.001).
Twelve of the children with CF were infected with P. aeruginosa at the time of the study. Comparison of lung function results by presence of P. aeruginosa infection is presented in Table 5. Children who were infected with P. aeruginosa had significantly higher LCI, but there was no significant difference in the other lung function measures between the two groups. The 18 children who were not infected with P. aeruginosa had significantly higher LCI (8.83 [1.61] versus 6.89 [0.44], mean difference 1.94 [1.13, 2.76], p < 0.001) and significantly higher sRaw z-score (1.59 [1.84] versus 0.00 [1.00], mean difference 1.59 [0.62, 2.56], p = 0.003) than the healthy children. There were no significant differences in spirometry results between these two groups.
This cross-sectional study compared MBW, body plethysmography, and spirometry results for preschool children with CF and healthy preschool children. Success rate for all measures was around 80%, with 65% of subjects completing all three tests on their first visit. Although group differences were seen with all three techniques, analysis of individual results showed that more children with CF had abnormal MBW results than had abnormal plethysmography or spirometry. One explanation for these results is that CF lung disease begins very early in life, and that this disease is not detectable by commonly used measures of lung function.
Previous Reports of Lung Function in Preschool Children With CF The poor success rate reported by Couriel and colleagues (19) contrasts markedly with our own experience in a younger population, where MBW results were obtained in 79% of children, including 48 (77%) of 62 children under 5 years of age. This difference in success rate may result from the modification to the MBW technique used in the current study. Couriel and colleagues (19) asked children to use a snorkel mouthpiece, and make no mention of distraction techniques. In the present study we used a facemask sealed with therapeutic putty, and measured children in a darkened room while they were watching a children's entertainment video of their choice. By this means it was possible to settle children and obtain recordings without leak. The measurement of sRaw by body plethysmography has been previously described in this age group (21–24). These previous reports have described measurement of sRaw using a modified facemask, and with children sometimes accompanied by an adult in the plethysmograph. In the current study children were asked to use a mouthpiece and to sit alone in the plethysmograph. The success rate for sRaw at first visit (78%) is nonetheless similar to that reported by other groups. Nielsen and colleagues (8) have used sRaw as an outcome measure in one longitudinal study of young children with CF. In this study, 30 children with CF (of whom 13 were aged 6 years or less at the start of the study) were repeatedly tested over a 4-year period. Seventeen (57%) of the 30 children in Nielsen's study had raised sRaw at commencement, whereas none had raised Rint, only one had raised respiratory reactance, and only two had raised respiratory resistance (both the latter outcomes obtained from the impulse oscillation technique). In contrast to the other outcome measures, group mean sRaw was persistently abnormal over the 4-year period. Three groups have recently published spirometry data from large populations of healthy children (17, 18, 25). Sixty (78%) of the 77 children in the current study were able to produce FEV0.5 at their first visit. This success rate is similar to those reported from these earlier studies. Marostica and colleagues (26) recently reported spirometry in 39 children with CF, aged 3 to 6 years (mean age 5.3 years). FEV1 was abnormal in 8 (24%) of 33 children, FVC abnormal in 7 (24%) of 29, and FEF25–75 abnormal in 4 (14%) of 29. In the current study, LCI was abnormal in 22 (73%) of 30 children, sRaw was abnormal in 14 children (47%), FEV0.5 was abnormal in 2 children (7%), and FEF25–75 was abnormal in 4 (13%). Direct comparison with results reported from other centers is not possible, because lung function results in children with CF are affected by center, era, and subject age. In this context, it is noted that the population presented here is slightly younger than those reported by Nielsen and colleagues (8) and Marostica and colleagues (26).
Selection of Outcome Measures for This Study
Strengths and Weaknesses of This Study It is noted that the success rates for the three airway function measures are very similar in children aged 3 years and older. The order of testing in this study was fixed, however, as MBW first, plethysmography second, and spirometry third. Direct comparison of relative feasibility for the three methods is therefore not possible. This order of testing was selected because deep inspirations are known to affect airway diameter during subsequent tidal breathing; and measurements of LCI and sRaw could have been altered if spirometry was performed earlier. Because spirometry is technique and effort dependent, it is possible that fatigue toward the end of the testing session could have resulted in erroneous FEV0.5 and FEF25–75 results in some children. We have previously demonstrated, however, that spirometry results obtained in healthy children undertaking our study protocol are very similar to those reported by Eigen and colleagues (17). For practical reasons it was not possible to standardize the interval between the child's morning airway clearance session and subsequent lung function testing. A study performed in older children with CF has demonstrated that LCI may be affected by airway clearance maneuvers (29). This subject warrants further investigation in younger children. We have presented limited data on the clinical condition of our subjects. Ideally we would have liked to compare lung function results with measures of infection and inflammation obtained from bronchoalveolar lavage, or with structural changes measured on CT. These investigations are not routinely performed in our clinics, and could not be justified for the purpose of this study. We were able to compare results by the presence of current P. aeruginosa infection, and demonstrated that infected children had significantly higher LCI than those who were not infected.
Clinical Significance of Results Abnormal spirometry results are recognized predictors of prognosis in subjects with CF, particularly in those with advanced disease (2–4). Although sRaw is not commonly used as an outcome measure in CF, it is used as a measure of airway function in other pathologic conditions (27). The clinical significance of abnormal LCI in subjects with CF is yet to be determined. Raised LCI results from differences in specific ventilation between parallel lung units. This can result from asymmetric narrowing of the airway lumen at branch points throughout the airway tree, which in turn may be caused by inflammation, scarring, or obstruction by mucus, or be secondary to changes in airway tone. Additionally, inhomogeneity may result from parenchymal changes in the subtended lung units, resulting in changes in compliance, and differing time-constants for filling and emptying. Our results clearly show that far more children have an abnormal LCI than any other parameter measured, and only one child had a normal LCI but an abnormality in another measurement. It is possible that this means that LCI is an oversensitive measure of lung function. From a clinical perspective, the most important consideration is whether impaired gas mixing represents early changes of CF lung disease, which subsequently progress, or whether impaired gas mixing is an epiphenomenon, present in most children with CF, but unconnected to future prognosis. This question can only be answered by longitudinal studies, tracking changes in LCI and forced expiratory parameters; and intervention studies, examining the effect of treatment on ventilation distribution. The finding of early structural changes on high-resolution CT in patients with apparently early disease and normal spirometry (30), however, lends plausibility to the view that LCI is indeed measuring something of biologic relevance. This view is further supported by the finding of higher LCI in children infected with P. aeruginosa, which is known to affect prognosis adversely in children with CF. No relationship was seen between age and any of the lung function parameters tested here. This is not a surprising finding, because CF is a heterogenous condition, and a 3-year-old subject is not necessarily expected to have milder lung disease than a child 2 years older. A previous study from our own laboratory has shown, however, that school-age children with CF have a mean LCI of 11.5 (28). This compares with a mean LCI of 9.6 in the preschool children with CF presented here. Previous MBW studies, mostly performed during the 1980s and early 1990s, have investigated ventilation distribution in school-age children or adults with CF. These studies have already shown that indices of ventilation inhomogeneity are raised in subjects with CF compared with healthy subjects (7), and that ventilation inhomogeneity is correlated to airway resistance, FEV1, and Schwachman score in subjects with CF (20, 31, 32). Furthermore, there is evidence that ventilation inhomogeneity decreases during treatment of an acute exacerbation in subjects with CF (33). There are two further points that suggest that MBW measurements may be of clinical value in CF. First, a number of centers have now reported on the use of MBW in infants (34–36), suggesting that this is one measure of airway function that may be applicable throughout life. Second, the LCI values obtained in healthy preschool children in this study are very similar to those reported in healthy school-age children (7). Interpretation of other measures of airway function, such as those obtained from forced expiration, can be confounded by the effects of growth and development, complicating interpretation of data from subjects of different ages. If LCI is constant in healthy children of differing ages, interpretation of longitudinal data is facilitated. In conclusion, this study has demonstrated that abnormal ventilation distribution, as evidenced by an abnormally raised LCI, is found in most preschool children with CF, including many with normal spirometry measurements and normal sRaw. These findings suggest that destructive processes in the airways of children with CF may start early in life, and that these changes are detectable by MBW, but not always by more commonly used measures of lung function. MBW may have considerable potential for early detection of CF lung disease in preschool children.
The authors thank the children and families who took part in this study. They also thank Ms. Clare Saunders, Portex Respiratory Unit, Institute of Child Health, who assisted with some of the measurements, and Dr. Angela Wade, Department of Epidemiology and Biostatistics, Institute of Child Health, who provided advice on data analysis. They also thank Dr. Robert Dinwiddie for scoring chest radiographs (data presented in online supplement) and Ms. Ammani Prasad for her assistance in collating these data.
Supported by Dunhill Medical Trust, British Lung Foundation, Cystic Fibrosis Trust, Portex Ltd. Research at the Institute of Child Health and Great Ormond Street Hospital for Children NHS Trust benefits from R&D funding received from the NHS Executive. This article has an online supplement, which is accessible from this issue's table of contents at www.atsjournals.org Conflict of Interest Statement: P.A. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; A.B. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; P.G. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; C.O. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; C.W. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; J.P. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; J.S. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; S.C. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; J.S. received an educational grant of $35,000 over two years (2001–2003) for development of training material for infant lung function testing from VIASYS. Received in original form July 12, 2004; accepted in final form October 25, 2004
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
METHODS
