LAUSR

In 2001, 5 of 16 (31%) children with CF and normal spirometry were found to have radiological bronchiectasis and in another group of children, in 2004, bronchiectasis was found in (11 of 37) 30% of children with CF and normal spirometry. Since that time bronchiectasis has been found in children with CF within the first months of life. These findings have suggested that CT scanning, and particularly the identification of bronchiectasis, may be a useful imaging outcome measure for early CF lung disease. All published CT scoring systems for CF include an assessment of bronchiectasis. Bronchiectasis is the most heavily weighted observation in the Brody or CFCT score, and it is at the top of the hierarchy of observations used to calculate the PerthRotterdam Annotated Grid Morphometric Analysis (PRAGMA) score. The presence of bronchiectasis on CT scanning is determined by assessing the size of the airway relative to the size of the accompanying pulmonary artery branch, termed the bronchoarterial ratio. This takes advantage of the similar tapering of the airway and the vascular trees to provide an internal reference for bronchial size without the need to define the exact location along the length of the bronchus. Initially proposed in the first years of the use of CT scanning in the lung, a bronchoarterial ratio of greater than 1:1 (some authors suggest taking >1.5) was suggested as the definition of bronchiectasis. It is unclearwho first proposed this ratio, but it is clear that it was based on studies of established bronchiectasis in adults, using the CT technology of the 1980s. This has arguable limited application to children although this measurement is widely used. In addition, its accuracy is limited by at least two sources of error. The first arises because of an incomplete definition of the bronchoarterial ratio. The arterial wall cannot be distinguished from the blood-filled lumen, so there is only a single possible measurement of vessel diameter. The bronchus, however, is air filled and so bronchial size can be assessed by measuring the luminal diameter, or the outer wall diameter. When specified, the luminal diameter has been more commonly, but not universally, used. In In one study of adults without cardiopulmonary disease, the bronchial wall thickness was 20% of the outer diameter of the bronchus, so the outer wall diameter in a normal airway can be expected to be in the range of 50% greater than the luminal diameter. The upper limit of normal in airway size would therefore be 50% greater if the luminal, rather than the outer wall, diameter was used to determine the bronchoarterterial ratio. The second source of error is the choice of a 1:1 ratio as the upper limit of normal. The bronchoarterial ratio is fairly constant in different parts of the lung, but increases with age and is well below 1:1 in healthy subjects. Thus, taking the same single cut-off for this ratio for both young children and adults is erroneous. A study involving adults (aged >21 years) without any cardiopulmonary disease found that a significant high correlation (r = 0.768, P < 0.0001) between age and the bronchoarterial ratio. There have been few studies of the bronchoarterial ratio in young children without lung disease. Indeed, we are aware of only 2 such published studies. In Ohio, Long and colleagues measured luminal diameter and found a ratio of approximately 0.49:1 in 20 children without respiratory disease, and mean+ 2SD of approximately 0.75. Although, unplanned the second scorer in this study measured the outer diameter and the corresponding values were mean ratio of 0.58 (plus 2SD = 0.78). In a Brisbane (Australia) study, the ratio for the cohort of 41 children without respiratory disease was 0.63 (95%CI, 0.61-0.65) and mean plus 2 SD was 0.76. In both these studies ‘well’ children did not have any respiratory disease unlike Kuo and colleagues’ study where the indication for undertaking CT scan in 10 of the 12 ‘disease controls’ was to evaluate whether bronchiectasis was present. CT scanning is now being used to assess the lungs in infants with CF. Unlike the assessment of the progression of established bronchiectasis, the goal is now to detect the earliest signs of bronchiectasis. Use of a single standard for the measurement of the bronchoarterial ratio will remove an important source of variability between different studies, whether comparing different populations or different scoring systems. This information should be specified in all studies. Readers should be aware of this potential source of errorwhen comparing studies, although given that it is standard to measure the inner luminal diameter, one can only assume that most studies would have used this diameter. Does use of the outer diameter as proposed by Kuo and colleagues add advantage to the long held traditional method of using the inner diameter? As calculated above, measuring the outer wall would reduce the upper limit of normal by about 1/3 which more closely approximates the upper limits of normal found by Long et al and by Kapur et al. There is far too little data at this point to determine how using the outer diameter would affect either research or clinical care, and hence the current standard of using the inner diameter remains the gold standard. However, the paper reminds the pediatric community of the need to develop children-specific data and not adopt adult criteria where it had been clearly shown that the BA ratio in those