Comments
Abstract
Rationale: Chest computed tomography (CT) is the gold standard for demonstrating cystic fibrosis (CF) airway disease. However, there are no standardized outcome measures appropriate for children younger than 6 years.
Objectives: We developed the Perth-Rotterdam Annotated Grid Morphometric Analysis for CF (PRAGMA-CF), a quantitative measure of airway disease, and compared it with the commonly used CF-CT scoring method.
Methods: CT scans from the Australian Respiratory Early Surveillance Team for CF (AREST CF) cohort in Western Australia were included. PRAGMA-CF was performed by annotating a grid overlaid on 10 axial slices for the presence of bronchiectasis, mucous plugging, or other airway abnormalities (inspiratory scans) and trapped air (expiratory scans). The separate proportions of total disease (%Dis), bronchiectasis (%Bx), and trapped air (%TA) were determined. Thirty scans were used for observer reliability, and 30 paired scans obtained at 1 and 3 years old were used for comparison with a validated standard and biologic plausibility.
Measurements and Main Results: Intraobserver, intraclass correlation coefficients (95% confidence interval) for %Dis, %Bx, and %TA were 0.93 (0.86–0.97), 0.93 (0.85–0.96), and 0.96 (0.91–0.98), respectively. The change in %Dis (P = 0.004) and %Bx (P = 0.001) with PRAGMA-CF was related to neutrophil elastase presence at age 3, whereas only the change in bronchiectasis score was related to neutrophil elastase (P < 0.001) with CF-CT. Sample-size calculations for various effect sizes are presented.
Conclusions: PRAGMA-CF is a sensitive and reproducible outcome measure for assessing the extent of lung disease in very young children with CF.
Currently there are no standardized outcome measures appropriate for very young children with cystic fibrosis (CF). This effectively excludes infants and young children from clinical trials, in an era where new, potentially disease-modifying drugs are becoming available. Chest computed tomography (CT) is the gold standard for detecting early CF lung disease, so a validated outcome measure from CT is urgently needed.
This study introduces PRAGMA-CF, a new quantitative outcome measure for chest CT appropriate for young children with early CF lung disease. It shows that PRAGMA-CF is repeatable, sensitive to early changes, and feasible for use as a clinical trial outcome.
Structural lung disease (SLD) in cystic fibrosis (CF) begins early in life, is progressive, and is often the only evidence of respiratory disease in children younger than 6 years of age (1–4). Chest computed tomography (CT) is the gold standard for demonstrating CF-related SLD. However, there are no validated quantitative outcomes based on CT that are appropriate for young children with early lung disease. Therefore, in an era that promises new disease-modifying therapies, an accurate and sensitive assessment tool is urgently needed for monitoring disease progress and for clinical trials.
Several CT scoring systems for CF have been developed for adults and children over the age of 6, with CF-CT being the most comprehensively examined (5–8). These methodologies are semiquantitative, and have not been designed for assessing subtle appearances and low extents of structural changes found early in life (3). In particular, the extent scores for components of SLD are generally based on the approximate proportion of the lung affected, with the smallest increment being one-third to one-half of a lobe (3, 5). Most young children have isolated disease, only the mildest severity of lung abnormalities, and lobar disease extent that is well below 50% (9–11). This creates, using currently available scoring systems, a binary outcome measure for each lobe in most scans. Therefore, a quantitative measure that is sensitive to early SLD is essential for use in clinical trials or longitudinal assessment in this young age group.
This paper introduces the Perth-Rotterdam Annotated Grid Morphometric Analysis method (PRAGMA-CF), a novel quantitative measure of CT SLD that specifically targets early CF lung disease. We assessed the intraobserver and interobserver agreement for PRAGMA-CF, relations with factors known to be associated with the development and progression of SLD, and compared the strengths of the associations using both PRAGMA-CF and CF-CT methods.
We hypothesized that during the first 3 years of life, the PRAGMA-CF approach compared with the CF-CT algorithm will have better repeatability, be more sensitive to disease progression, and more strongly correlate with airway inflammatory biomarkers. Some of the results of these studies have been previously reported in the form of abstracts (12–14).
The children in this study were participants in the Australian Respiratory Early Surveillance Team for Cystic Fibrosis (AREST CF) early surveillance program. To determine the minimum number of annotated slices required for accurate analysis, we selected 20 volumetric inspiratory and expiratory scans representing a range of SLD extents (setslice): 10 scans from patients aged younger than 6 years with the highest CF-CT bronchiectasis score (“severe” cases), and 10 scans randomly selected from patients aged 3 and younger (“mild” cases). To determine intraobserver and interobserver repeatability, 30 additional volumetric inspiratory and expiratory scans were randomly selected from patients aged younger than 6 years (setrep). Random scan selection was performed by assigning all eligible CTs scan a pseudorandom number in the interval (−1 to 1), sorting the list of scans based on this value, and selecting the required number of scans with the lowest rank (ensuring that the same patient was not included more than once).
To assess biologic plausibility and assess PRAGMA-CF as a potential clinical trial outcome for infants, we examined data from all patients who had paired volumetric inspiratory CT scans (and either volumetric or limited slice expiratory scans) acquired at both the ages of 1 and 3 years (setbio).
Chest CT scans were obtained under general anesthesia as previously described (3).
CT scans were scored using a simplified version of the CF-CT scoring system (9, 15) and the PRAGMA-CF algorithm.
We developed PRAGMA-CF based on the severe advanced lung disease method described by Loeve and colleagues (16). Using the thinnest slice reconstruction (0.8–1.0 mm) of the inspiratory scan, 10 slices equally spaced between the apex and the base of the lung were selected for annotation and overlaid with a square grid. We corrected for lung size by setting the grid cell size equal to one-twentieth of the lung width at the carina (rounded to the nearest mm), measured with electronic calipers using Myrian software (Intrasense, Montpellier, France). This size was chosen because it approximately corresponds to the size of the largest airways.
For each grid cell that was at least 50% covered by the lung field, any presence of airway disease was annotated according to a hierarchical system, from highest to lowest priority: bronchiectasis (outer edge bronchus-artery cross-sectional area ratio > 1), mucous plugging (high-density airway occlusion or tree-in-bud appearance), bronchial wall thickening (airway walls that are thicker or have increased signal intensity relative to normal airways, assessed subjectively), atelectasis, or normal lung structure (Figures 1A and 1B). Although only select slices were annotated, the observer was permitted to scroll through the entire lung volume to assist in classification, for example to distinguish between an occluded airway and an artery. Cells containing atelectasis were annotated but excluded from all analysis because they are likely related to general anesthesia rather than pathology (17). The primary outcomes from this method are the volume proportion of the lung with airway disease (%Dis), determined by dividing the number of cells annotated with either bronchiectasis, mucous plugging, or airway abnormality by the total number of annotated cells (excluding atelectasis), and the volume proportion of the lung with bronchiectasis (%Bx), determined by dividing the number of cells annotated with bronchiectasis by the total number of annotated cells (excluding atelectasis).
Figure 1. (A) Slice from inspiratory computed tomography scan with overlaid grid. (B) Slice annotated with PRAGMA-CF (inspiratory): red = bronchiectasis; yellow = mucous plugging; orange = bronchial wall thickening; pink = atelectasis; green = no abnormality seen. (C) Minimum intensity projection filtered slice from expiratory computed tomography scan with overlaid grid. (D) Slice annotated with PRAGMA-CF (expiratory): blue = trapped air; green = not trapped air.
[More] [Minimize]Trapped air was assessed with a similar methodology on the expiratory scans. A minimum intensity projection (MinIP) reconstruction was performed before analysis with a slice thickness of 4–5 mm, depending on the clinical protocol at the time. This projection increases visibility of low-intensity regions and facilitates visualization of trapped air (18, 19). Cells were annotated according to whether trapped air represented 50% or more of the lung field in the cell (trapped air) or less than 50% (healthy). The primary outcome (Figures 1C and 1D) was the volume proportion of the lung with trapped air (%TA), determined by dividing the number of trapped air cells by the total number of annotated cells.
Before 2010, clinical procedure was to acquire expiratory scans using a three-slice protocol. For these scans, the lung width was measured on the second (middle) slice, and all three slices were annotated for the presence of trapped air. Because MinIP reconstructions are not possible on three-slice scans, standard 1-mm slice thickness lung reconstructions were used for all expiratory scans in setbio to minimize the effect of scan technique on longitudinal assessment of trapped air.
Scans from setslice were assessed using PRAGMA-CF, with slices annotated at 5-mm intervals to determine the most accurate estimate of %Dis, %Bx, and %TA. Every second annotated slice was then removed from analysis, representing PRAGMA-CF at 10-mm intervals, and then again representing PRAGMA-CF at 20-mm intervals. The maximum distance between annotated slices was determined by finding the largest interval that did not result in a significant change in PRAGMA-CF outcomes. Based on these results, a fixed number of 10 annotated slices was chosen as the standard for PRAGMA-CF. This standard was used for assessing all other data sets, and takes approximately 20 minutes to analyze.
The primary observer for this study (T.R.) was a doctoral student, trained in CF-CT scoring, with more than 3 years of experience in early CF lung disease. A second observer (M.C.J.O.), a medical student previously trained in CF-CT scoring, was provided with a detailed instruction manual on PRAGMA-CF and a practice set of six batches of five CT scans to annotate. Detailed feedback based on visual assessment of the annotations was given after each batch.
To determine interobserver reliability, scans from setrep, were deidentified and scored in random order with PRAGMA-CF independently by both observers. This set was rerandomized and rescored after 1 month (to reduce recall bias) by the primary observer to determine intraobserver repeatability.
We reasoned that any new assessment of SLD should reflect known pathobiologic factors. We therefore compared relations between inflammatory biomarkers and both the PRAGMA-CF and CT-CT outcomes. Following the CT scan, bronchoalveolar lavage was performed to determine IL-8 concentration, neutrophil count, and neutrophil elastase (NE) presence, as previously described (2, 20).
Slice annotation interval scores were analyzed pairwise by spacing, using the intraclass correlation coefficient (ICC). Interobserver and intraobserver reliability was assessed using the ICC (two-way mixed effects model) and Bland-Altman analysis (21). In general, ICC between 0.4 and 0.6 is considered moderate, between 0.6 and 0.8 good, and greater than 0.8 excellent (22). CF-CT scores were treated as continuous variables for all analyses. PRAGMA-CF and CF-CT scores were compared against binary and continuous outcomes using the Wilcoxon rank sum test and Spearman rank correlation (ρs), respectively.
All analyses were performed using Stata version 13.0 (Stata Corp., College Station, TX).
Median (range) PRAGMA-CF outcomes for setslice for %Dis, %Bx, and %TA at 5-mm intervals were 3.54 (0.00–23.76), 1.76 (0.00–16.91), and 2.05 (0.00–51.51), respectively. ICCs between 5- and 10-mm intervals and between 5- and 20-mm intervals exceeded 0.99 for all outcome measures (see online supplement). Ten slices was chosen as the standard for PRAGMA-CF, so that the volume of lung annotated is standardized across lung heights, and thus ensuring that annotation intervals are less than 20 mm for this age group.
Median (range) PRAGMA-CF outcomes for setrep for %Dis, %Bx, and %TA were 3.99 (0.72–11.59), 0.00 (0.00–6.84), and 1.52 (0.00–55.36), respectively. The ICC was greater than 0.8 for all outcomes, both for intraobserver and interobserver reliability (Table 1).
| ICC (95% CI) | Mean Difference* | |
|---|---|---|
| Intraobserver | ||
| %Dis | 0.929 (0.857–0.966) | n/s |
| %Bx | 0.928 (0.854–0.965) | 0.32 (P = 0.005) |
| %TA | 0.960 (0.916–0.981) | 2.03 (P = 0.001) |
| Interobserver | ||
| %Dis | 0.851 (0.710–0.926) | 0.78 (P = 0.006) |
| %Bx | 0.907 (0.815–0.955) | −0.38 (P = 0.006) |
| %TA | 0.938 (0.873–0.970) | n/s |
Thirty patients had paired volumetric inspiratory scans at the ages of 1 and 3 (Table 2). Of these, six patients had volumetric expiratory scans at age 1, and 25 at age 3; the remainder of expiratory scans were three-slice. A change in expiratory scan methodology (volumetric vs. three-slice) between visits was not related to the change in %TA (P = 0.576).
| Clinical characteristics | |
| N | 30 |
| Sex, M/F | 10:20 |
| Severe genotype*† | 22/27 (81%) |
| Pancreatic sufficient† | 7 (23%) |
| Respiratory symptoms† | |
| Age 1 | 6 (20%) |
| Age 3 | 7 (25%) |
| PRAGMA-CF | |
| %Disease | |
| Age 1 | 0.79 (0.00 to 4.88) |
| Age 3 | 1.86 (0.00 to 10.53) |
| Change | 0.82 (−1.06 to 5.64) |
| %Bronchiectasis | |
| Age 1 | 0.00 (0.00 to 0.62) |
| Age 3 | 0.00 (0.00 to 6.72) |
| Change | 0.00 (−0.22 to 6.36) |
| %Trapped air | |
| Age 1 | 4.21 (0.00 to 29.79) |
| Age 3 | 1.32 (0.00 to 59.79) |
| Change | −1.63 (−20.29 to 47.34) |
| CF-CT scores | |
| Total | |
| Age 1 | 4 (0 to 19) |
| Age 3 | 11 (0 to 29) |
| Change | 7 (−1 to 16) |
| Bronchiectasis | |
| Age 1 | 0 (0 to 5) |
| Age 3 | 1 (0 to 10) |
| Change | 1 (0 to 8) |
| Trapped air | |
| Age 1 | 2 (0 to 5) |
| Age 3 | 1 (0 to 11) |
| Change | 0 (−5 to 10) |
| Inflammation | |
| Neutrophils | |
| Age 1 | 1.06 (0.03 to 31.61) |
| Age 3 | 4.03 (0.00 to 223) |
| IL-8 | |
| Age 1 | 0.48 (0.05 to 5.84) |
| Age 3 | 1.38 (0.05 to 30.95) |
| NE present† | |
| Age 1 | 2/30 (7%) |
| Age 3 | 6/30 (20%) |
There were significant cross-sectional relationships between all inspiratory CT outcomes and inflammatory markers (Table 3). For expiratory scans, %TA was significantly related to neutrophil count and NE presence, whereas no relationships were found with CF-CT trapped air score. Longitudinal change in %Dis was significantly related to neutrophil count and NE presence at age 3; however, total CF-CT score was related only to IL-8 concentration. The change in both %Bx and CF-CT bronchiectasis score were significantly related to all inflammatory markers at age 3 (Table 4). Additional graphical data are provided in the online supplement.
| IL-8 Concentration | Neutrophil Count | NE Presence | |
|---|---|---|---|
| PRAGMA-CF | |||
| %Disease | |||
| Age 1 | 0.39 (0.035)* | 0.37 (0.045)* | 0.025* |
| Age 3 | 0.39 (0.036)* | 0.44 (0.015)* | 0.002* |
| Combined | 0.40 (0.002)* | 0.38 (0.003)* | <0.001* |
| %Bronchiectasis | |||
| Age 1 | 0.21 (0.264) | 0.15 (0.426) | 0.014* |
| Age 3 | 0.48 (0.069) | 0.66 (<0.001)* | <0.001* |
| Combined | 0.45 (<0.001)* | 0.48 (0.001)* | <0.001* |
| %Trapped air | |||
| Age 1 | 0.09 (0.622) | 0.38 (0.041)* | 0.243 |
| Age 3 | 0.27 (0.142) | 0.39 (0.035)* | 0.032* |
| Combined | 0.12 (0.367) | 0.36 (0.005)* | 0.017* |
| CF-CT scores | |||
| Total | |||
| Age 1 | 0.38 (0.040)* | 0.28 (0.137) | 0.153 |
| Age 3 | 0.54 (<0.001)* | 0.52 (0.003)* | 0.002* |
| Combined | 0.55 (<0.001)* | 0.40 (0.002)* | <0.001* |
| Bronchiectasis | |||
| Age 1 | 0.20 (0.295) | 0.15 (0.444) | 0.124 |
| Age 3 | 0.41 (0.023)* | 0.63 (<0.001)* | 0.001* |
| Combined | 0.41 (0.001)* | 0.43 (0.001)* | <0.001* |
| Trapped air | |||
| Age 1 | −0.03 (0.865) | −0.12 (0.518) | 0.232 |
| Age 3 | 0.36 (0.052) | 0.21 (0.260) | 0.651 |
| Combined | 0.18 (0.168) | 0.06 (0.643) | 0.938 |
| IL-8 Concentration | Neutrophil Count | NE Presence | |
|---|---|---|---|
| PRAGMA-CF | |||
| Δ%Disease | 0.18 (0.34) | 0.41 (0.025)* | 0.0043* |
| Δ%Bronchiectasis | 0.41 (0.026)* | 0.63 (0.0002)* | 0.001* |
| Δ%Trapped air | 0.06 (0.74) | 0.18 (0.33) | 0.058 |
| CF-CT scores | |||
| ΔTotal | 0.37 (0.047)* | 0.31 (0.096) | 0.11 |
| ΔBronchiectasis | 0.43 (0.016)* | 0.59 (0.0006)* | 0.0008* |
| ΔTrapped air | 0.30 (0.10) | 0.32 (0.09) | 0.64 |
%Dis at age 1 was significantly related to the change in %Bx (ρs = 0.44; P = 0.016). This relationship was also present between total CF-CT score at age 1 and the change in CF-CT bronchiectasis score (ρs = 0.36; P = 0.050). No other CT outcomes at age 1 were related to bronchiectasis progression.
There were significant correlations between %TA and %Dis (ρs = 0.50; P < 0.001), and %TA and %Bx (ρs = 0.42; P < 0.001). These relationships were not significant between CF-CT trapped air score and total score (ρs = 0.22; P = 0.091) or bronchiectasis score (ρs = 0.17; P = 0.197). Longitudinally, the change in %TA was significantly related to the change in %Dis (ρs = 0.62; P < 0.001) and the change in %Bx (ρs = 0.47; P = 0.008). These relationships were not significant between changes in CF-CT trapped air score and total score (ρs = 0.12; P = 0.533) or bronchiectasis score (ρs = 0.30; P = 0.107).
We performed power calculations based on a range of effect sizes. The mean (SD) %Dis at 3 years was 2.62 (2.59); we calculated sample sizes based on trials aiming to reduce this by 75, 50, and 25%. In addition, 46% of patients either developed or had progression in the extent of bronchiectasis; we also calculated sample sizes aimed at reducing this proportion to 10, 20, and 30% (Table 5).
| Sample Size | |
|---|---|
| Relative reduction in %Dis at age 3 | |
| 75% | 46 |
| 50% | 100 |
| 25% | 390 |
| Proportion with %Bx progression | |
| 10% | 36 |
| 20% | 76 |
| 30% | 208 |
This study presents the first quantitative method for assessing SLD on chest CT designed specifically for young children and infants with CF. We confirmed that although the prevalence of SLD early in life is high (1–3), the extent is relatively low, with a median proportion of lung volume affected being 0.87 and 1.86% at age 1 and 3, respectively. The most important findings of this study are that the quantitative PRAGMA-CF outcomes (1) have high intraobserver and interobserver agreement, (2) are better correlated to neutrophilic inflammation than CF-CT scores, and (3) show stronger relationships between structural changes and trapped air progression.
ICCs were all greater than 0.90 for intraobserver and greater than 0.85 for interobserver comparisons, similar to previously published intraobserver repeatability for simplified CF-CT scores of bronchiectasis and trapped air (9). The mean differences between raters and between the first and second scoring attempts, although significant, were small in magnitude, and likely skewed by a few high-scoring patients. Although there is no direct comparison between CF-CT outcomes and %Dis, it is as repeatable as the other PRAGMA-CF outcomes. PRAGMA-CF therefore is at least as repeatable as CF-CT, but over a much more sensitive range (several percent of the lung compared with a binary per-lobe score of present/absent).
The number of patients with detectable NE was low, and therefore further studies are required to determine the true relationship between CT and NE. Cross-sectional relationships between inflammatory markers were similar for PRAGMA-CF and CF-CT scores, with the exception of trapped air, where PRAGMA-CF was better correlated. Longitudinally, the change in %Dis was related to the presence of NE and the neutrophil count at age 3, whereas the change in total CF-CT score was only related to IL-8. Because CF-CT is not quantitative, it is not clear how much of the total score relationship is driven by the bronchiectasis component. The change in both %Bx and the CF-CT bronchiectasis score was significantly related to all inflammatory markers at age 3 (Table 4). This suggests that PRAGMA-CF is more sensitive than CF-CT–derived scores in assessing early changes in lung structure caused by inflammation.
%Dis at the age of 1 was significantly related to the change in %Bx over the 2 years, suggesting that patients with worse baseline disease have faster progression of bronchiectasis. Therefore, PRAGMA-CF performed at the age of 1 year can potentially be used to identify patients at high risk for disease progression.
Trapped air assessed using PRAGMA-CF was significantly related to other CT outcomes (both cross-sectionally and longitudinally), but not when assessed with CF-CT scores. Although the relationship between trapped air and bronchiectasis progression has been previously demonstrated in children (1, 23), the progression of SLD in general and trapped air in children aged younger than 3 years has not been observed. Our data suggest that this is because of the semiquantitative nature of CF-CT and related scoring systems, rather than pathobiologic reasons.
These data suggest that PRAGMA-CF could be an appropriate clinical trial outcome in young children, particularly for therapies aimed at reducing early inflammatory SLD. Because clinical trials have not previously been available for infants, the clinical impact of SLD extent in infancy is unknown. However, the sample sizes presented for a range of potential effect sizes suggest that multicenter trials in children between 1 and 3 years run over 1–2 years are feasible, with PRAGMA-CF as an outcome.
Grid cell size was arbitrarily chosen to be one-twentieth of the lung width. This size was selected because it approximately represents the size of the largest assessable airway in the lung. Although objective comparisons between grid resolutions could determine the optimum cell size, the current value is already sufficient to demonstrate the use of PRAGMA-CF. Future studies should be undertaken to assess the repeatability of PRAGMA-CF in a variety of age groups and disease severities, because it is possible that patients with mild disease or of young age, as in this study, may have a larger relative variability of PRAGMA-CF outcomes compared with older subjects with more severe disease. PRAGMA-CF lends itself to automations, such as by using textural analysis (24), which mitigate observer variability. However, even without further development, significant relationships between %Dis at age 1 and inflammation and bronchiectasis at age 3 were present, suggesting that PRAGMA-CF results at 1 year are likely to be clinically relevant and therefore relevant as a trial end point.
A limitation of the study with regard to trapped air assessments was that some expiratory scans were performed using a three-slice protocol. Because we found no statistical difference in %TA in patients who had the same expiratory scan method on both visits, as compared with those who had a different method, we pooled the data for biologic validation. Another limitation is that MinIP reconstructions were not used for biologic validation. Because MinIP reconstructions improve visualization of low-intensity regions, and hence reduce observer variability, we are likely to have underestimated any relationships between disease markers and PRAGMA-CF trapped air measurements.
The CF-CT scoring system has been used to evaluate SLD in young children with CF (3, 25–27), but is relatively insensitive to mild disease. Our data using PRAGMA-CF demonstrate that reliable quantitative estimates of lung disease can be obtained in young children. These observations are significant because young children may stand to benefit the most from new, disease-modifying drugs that are becoming available, where until now, there has been a lack of a suitable outcome measure to enable clinical trials (15). Our data suggest that PRAGMA-CF is a promising candidate as a standardized outcome measure for use in intervention studies and clinical trials in young children.
We believe that the circumstances now exist to make CT a rational choice as a primary outcome measure in clinical trials involving very young children with CF. Technologic advances have resulted in improvements in CT image quality for a given radiation dose and improvements are likely to continue (28). In an intervention study that requires images at two time points, the estimated excess risk of cancer is so low as to be incalculable (29–31). Furthermore, using PRAGMA-CF, the sample sizes required for an intervention study with a relatively short duration (i.e., 2 yr), and the ubiquitous access at CF centers to CT scanners, mean that multiple studies can be undertaken simultaneously around the world. This is an important consideration given the low incidence of CF and the exciting prospects that now exist with regard to the development of disease-modifying interventions.
The authors thank the Department of Diagnostic Imaging at Princess Margaret Hospital for Children for assistance with the acquisition and collation of computed tomography scans, and Marcel Koek from Erasmus MC for the development of the grid software used for computed tomography annotation.
| 1. | Mott LS, Park J, Murray CP, Gangell CL, de Klerk NH, Robinson PJ, Robertson CF, Ranganathan SC, Sly PD, Stick SM; AREST CF. Progression of early structural lung disease in young children with cystic fibrosis assessed using CT. Thorax 2012;67:509–516. |
| 2. | Sly PD, Brennan S, Gangell C, de Klerk N, Murray C, Mott L, Stick SM, Robinson PJ, Robertson CF, Ranganathan SC; Australian Respiratory Early Surveillance Team for Cystic Fibrosis (AREST-CF). Lung disease at diagnosis in infants with cystic fibrosis detected by newborn screening. Am J Respir Crit Care Med 2009;180:146–152. |
| 3. | Stick SM, Brennan S, Murray C, Douglas T, von Ungern-Sternberg BS, Garratt LW, Gangell CL, De Klerk N, Linnane B, Ranganathan S, et al. Bronchiectasis in infants and preschool children diagnosed with cystic fibrosis after newborn screening. J Pediatr 2009;155:623–628, e1. |
| 4. | Sly PD, Gangell CL, Chen L, Ware RS, Ranganathan S, Mott LS, Murray CP, Stick SM; AREST CF Investigators. Risk factors for bronchiectasis in children with cystic fibrosis. N Engl J Med 2013;368:1963–1970. |
| 5. | Brody AS, Klein JS, Molina PL, Quan J, Bean JA, Wilmott RW. High-resolution computed tomography in young patients with cystic fibrosis: distribution of abnormalities and correlation with pulmonary function tests. J Pediatr 2004;145:32–38. |
| 6. | de Jong PA, Ottink MD, Robben SG, Lequin MH, Hop WC, Hendriks JJ, Paré PD, Tiddens HA. Pulmonary disease assessment in cystic fibrosis: comparison of CT scoring systems and value of bronchial and arterial dimension measurements. Radiology 2004;231:434–439. |
| 7. | Loeve M, Gerbrands K, Hop WC, Rosenfeld M, Hartmann IC, Tiddens HA. Bronchiectasis and pulmonary exacerbations in children and young adults with cystic fibrosis. Chest 2011;140:178–185. |
| 8. | Robinson TE, Leung AN, Chen X, Moss RB, Emond MJ. Cystic fibrosis HRCT scores correlate strongly with Pseudomonas infection. Pediatr Pulmonol 2009;44:1107–1117. |
| 9. | Mott LS, Park J, Gangell CL, de Klerk NH, Sly PD, Murray CP, Stick SM; Australian Respiratory Early Surveillance Team for Cystic Fibrosis Study Group. Distribution of early structural lung changes due to cystic fibrosis detected with chest computed tomography. J Pediatr 2013;163:243–248, e1–e3. |
| 10. | Wainwright CE, Vidmar S, Armstrong DS, Byrnes CA, Carlin JB, Cheney J, Cooper PJ, Grimwood K, Moodie M, Robertson CF, et al.; ACFBAL Study Investigators. Effect of bronchoalveolar lavage-directed therapy on Pseudomonas aeruginosa infection and structural lung injury in children with cystic fibrosis: a randomized trial. JAMA 2011;306:163–171. |
| 11. | Thia LP, Calder A, Stocks J, Bush A, Owens CM, Wallis C, Young C, Sullivan Y, Wade A, McEwan A, et al.; London Cystic Fibrosis Collaboration. Is chest CT useful in newborn screened infants with cystic fibrosis at 1 year of age? Thorax 2014;69:320–327. |
| 12. | Rosenow T, Tiddens H, de Bruijne M, Berry L, Stick S. Quantitation of chest CT abnormalities in early life CF: back to the basics. In: The 27th Annual North American Cystic Fibrosis Conference. Hoboken, NJ: Wiley-Blackwell; 2013. pp. 340–340. |
| 13. | Rosenow T, Tiddens H, De Bruijne M, Stick S. PRAGMA: a new method of quantifying structural lung disease in young children with cystic fibrosis. In: Thoracic Society of Australia and New Zealand Annual Scientific Meeting. Hoboken, NJ: Wiley-Blackwell; 2014. p. 48. |
| 14. | Rosenow T, Tiddens H, Oudraad M, de Bruijne M, Stick S. PRAGMA: further support for use as a quantitative CT outcome measure. In: The 28th Annual North American Cystic Fibrosis Conference. Hoboken, NJ: Wiley-Blackwell; 2014. p. 295. |
| 15. | Stick SM. Clinical trials in infants with cystic fibrosis detected following newborn screening. Paediatr Respir Rev 2008;9:176–180. |
| 16. | Loeve M, van Hal PT, Robinson P, de Jong PA, Lequin MH, Hop WC, Williams TJ, Nossent GD, Tiddens HA. The spectrum of structural abnormalities on CT scans from patients with CF with severe advanced lung disease. Thorax 2009;64:876–882. |
| 17. | Newman B, Krane EJ, Gawande R, Holmes TH, Robinson TE. Chest CT in children: anesthesia and atelectasis. Pediatr Radiol 2014;44:164–172. |
| 18. | Bhalla M, Naidich DP, McGuinness G, Gruden JF, Leitman BS, McCauley DI. Diffuse lung disease: assessment with helical CT—preliminary observations of the role of maximum and minimum intensity projection images. Radiology 1996;200:341–347. |
| 19. | Fotheringham T, Chabat F, Hansell DM, Wells AU, Desai SR, Gückel C, Padley SP, Gibson M, Yang GZ. A comparison of methods for enhancing the detection of areas of decreased attenuation on CT caused by airways disease. J Comput Assist Tomogr 1999;23:385–389. |
| 20. | Fuchs SI, Gappa M, Eder J, Unsinn KM, Steinkamp G, Ellemunter H. Tracking Lung Clearance Index and chest CT in mild cystic fibrosis lung disease over a period of three years. Respir Med 2014;108:865–874. |
| 21. | Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986;1:307–310. |
| 22. | Loeve M, de Bruijne M, Hartmann IC, van Straten M, Hop WC, Tiddens HA. Three-section expiratory CT: insufficient for trapped air assessment in patients with cystic fibrosis? Radiology 2012;262:969–976. |
| 23. | Tepper LA, Utens EM, Caudri D, Bos AC, Gonzalez-Graniel K, Duivenvoorden HJ, van der Wiel EC, Quittner AL, Tiddens HA. Impact of bronchiectasis and trapped air on quality of life and exacerbations in cystic fibrosis. Eur Respir J 2013;42:371–379. |
| 24. | Ciompi F, Palaioroutas A, Loeve M, Pujol O, Radeva P, Tiddens H, de Bruijne MLung tissue classification in severe advanced cystic fibrosis from CT scans. Presented at the Fourth International Workshop on Pulmonary Image Analysis. September 18, 2011, Toronto, Canada. pp. 57–68. |
| 25. | Brody AS, Kosorok MR, Li Z, Broderick LS, Foster JL, Laxova A, Bandla H, Farrell PM. Reproducibility of a scoring system for computed tomography scanning in cystic fibrosis. J Thorac Imaging 2006;21:14–21. |
| 26. | Calder AD, Bush A, Brody AS, Owens CM. Scoring of chest CT in children with cystic fibrosis: state of the art. Pediatr Radiol 2014;44:1496–1506. |
| 27. | Tiddens HA, Rosenow T. What did we learn from two decades of chest computed tomography in cystic fibrosis? Pediatr Radiol 2014;44:1490–1495. |
| 28. | Newell JD Jr, Fuld MK, Allmendinger T, Sieren JP, Chan KS, Guo J, Hoffman EA. Very low-dose (0.15 mGy) chest CT protocols using the COPDGene 2 test object and a third-generation dual-source CT scanner with corresponding third-generation iterative reconstruction software. Invest Radiol 2015;50:40–45. |
| 29. | Kuo W, Ciet P, Tiddens HA, Zhang W, Guillerman RP, van Straten M. Monitoring cystic fibrosis lung disease by computed tomography. Radiation risk in perspective. Am J Respir Crit Care Med 2014;189:1328–1336. |
| 30. | National Research Council, Committee to Assess Health Risks from Exposure to Low Levels of Ionizing Radiation. Health risks from exposure to low levels of ionizing radiation: BEIR VII-Phase 2. Washington, DC: National Research Council; 2005. |
| 31. | The American Association of Physicists in MedicineAAPM position statement on radiation risks from medical imaging procedures. 2011 [accessed 2014 Nov 18]. Available from: http://www.aapm.org/org/policies/details.asp?id=318&type=PP¤t=true |
Supported by the Cystic Fibrosis Foundation (United States; project grant support), National Health and Medical Research Council (Australia; project grant support and fellowships), and Cystic Fibrosis Australia (project grant support).
Author Contributions: Literature search, T.R. and H.A.W.M.T. Computed tomography analysis, T.R., C.P.M., and M.C.J.O. Data and statistical analysis, T.R., L.T., and S.M.S. Generation of figures, T.R. and L.T. Study conception and design, data interpretation, manuscript drafting and revision for important intellectual content, manuscript editing, accountability for all aspects of the work, and final approval of the submitted manuscript, all authors.
This article has an online supplement, which is accessible from this issue's table of contents at www.atsjournals.org
Originally Published in Press as DOI: 10.1164/rccm.201501-0061OC on March 10, 2015
Author disclosures are available with the text of this article at www.atsjournals.org.
