© 2008 American Thoracic Society doi: 10.1164/rccm.200801-028UP
Update in Pediatric Lung Disease 20071 Imperial School of Medicine at National Heart and Lung Institute, London, United Kingdom; and 2 Royal Brompton Hospital, London, United Kingdom Correspondence and requests for reprints should be addressed to Andrew Bush, M.D., F.R.C.P., Department of Paediatric Respiratory Medicine, Royal Brompton Hospital, Sydney Street, London SW3 6NP, UK. E-mail: a.bush{at}rbh.nthames.nhs.uk Previous updates (1, 2) have reviewed important articles in the Journal covering general research issues (3); lung growth and repair (4–9); pulmonary hypertension (10–12); neonatal lung disease and its long-term importance (13–23); important aspects of gene–environment interactions (24); and specific childhood diseases, including viral infections (25–28), complications of neuromuscular disease (29), and pediatric interstitial lung disease (30), setting them in the context of work published elsewhere. This update discusses new articles that take forward the fields of the developing lung; clinical studies spanning the age ranges, which shed light on the importance of early lung disease and the tracking of lung function; new diagnostic tests; and significant developments in pediatric interstitial lung disease. As with previous updates, the topics and space allocation reflects subjective personal interest, not necessarily the value of the manuscript. THE DEVELOPING LUNG Because fetal and early-life events strongly influence long-term lung health in adults and even in the elderly (31, 32), prenatal lung development is an increasingly important area of study. Alveolar development is a largely postnatal phenomenon, and indeed, very preterm babies may be born with few, if any, alveoli in the conventional sense. It is known from a number of animal studies that positive-pressure ventilation, parenteral steroids, and hyperoxia, insults to which very preterm babies may need to be exposed to save their lives, all adversely impact alveolar development (33–35). The twin and related challenges facing pediatric pulmonology are the prevention of adverse effects on the developing alveoli, and the regeneration of new alveoli to replace those that are irretrievably damaged. There are tantalizing hints of what is possible; even very extensive pneumatoceles, presumably with major areas of alveolar destruction, can heal completely (36). We need to know how to recapitulate this process in the context of the premature lung, or after later destructive viral infections, leading to, for example, Swyer-James-Macloud syndrome. Nitric oxide (NO) has involvement in multiple physiologic and pathophysiologic processes and is also being used therapeutically, including in the prevention of chronic lung disease of prematurity (37–40). However, NO may also produce harm by the generation of reactive nitrogen species. Auten and coworkers (41) hypothesized that inhaled ethyl nitrite (ENO), which has already been used to treat newborns with pulmonary hypertension (42), might offer superior protection to the effects of hyperoxia compared with NO, but with a better safety profile (43). They used a free-breathing newborn rat model exposed either to air or 95% oxygen, with salvage of the 95% oxygen–exposed rats using either NO or ENO. They found that both compounds partially attenuated the leukocyte infiltration caused by hyperoxia, but importantly, there was strikingly good preservation of alveolar structure in the ENO- but not in the NO-salvaged rats. Clearly, the studies need to be extended into a model of a preterm ventilated animal to see if the same effects can be seen, but if these striking findings are confirmed, and ENO continues to appear to be safe, it may well find its way into the neonatal nursery as an alveolar protective agent, and possibly also into treatment for other diseases in which hyperoxia may be one of the factors implicated in alveolar damage. NO is an ubiquitous molecule, and bombesin may be following in its wake. Previous work has shown that increased bombesin immunoreactivity is the hallmark of neuroendocrine cell hyperplasia of infancy (NEHI) (44), and it has been implicated in the development of immune responses in bronchopulmonary dysplasia (BPD) (22). There is evidence that bombesin may be implicated in the development of BPD (45, 46). Subramaniam and colleagues (47) used the preterm baboon model to further explore the role of bombesin in BPD. Baboons were delivered at 125 days and given as-required oxygen, a model similar to "new" BPD, and were also treated with a bombesin blocking antibody or placebo. The major pulmonary bombesin-like peptide (gastrin-releasing peptide) mRNA was up-regulated eightfold. There was a dramatic increase in pulmonary neuroendocrine-like cells. Oxygenation index correlated with urinary bombesin-like peptide levels. Early overproduction of bombesin predicted subsequent alveolarization defects. Blockade of bombesin by the antibody 2A11 improved alveolarization. Bombesin effects were complex, because the peptide promoted tubulogenesis of cultured microvascular endothelial cells, which might be believed to be beneficial, although some might dispute this (23), and the effect was abrogated by cocultured fetal lung mesenchymal cells. Thus, bombesin and related peptides are an intriguing target in BPD, but we need more mechanistic data to unpick their protean effects before they become a realistic therapeutic target. Two articles have reported on aspects of abnormal lung development (48, 49). The selective serotonin reuptake inhibitor fluoxetine reverses pulmonary hypertension in the rat (50), but has been associated with persistent pulmonary hypertension of the newborn in infants whose mothers ingested it in pregnancy (51). Pregnant rats were treated with either fluoxetine or placebo, and the fetuses delivered by cesarean section (48). Fluoxetine led to fetal pulmonary hypertension as shown by increased right ventricular weight and pulmonary arterial medial thickness, and there was significantly increased smooth muscle cell proliferation. Fetal, but not adult, smooth muscle cells contracted in response to fluoxetine in vitro. This study is also important in reminding us of the developmental effects leading to age-related different responses to medications, underscoring the need not to rely on adult studies of medicines to guide pediatric pharmacotherapy, a view taken by the U.S. Food and Drug Administration and, more recently, the European Union (52). Unilateral congenital diaphragmatic hernia (CDH) is associated with bilaterally abnormal airway branching and alveogenesis. Boucherat and coworkers (49) performed a study of lung fibroblast growth factor 18 (FGF18) and elastin in human and animal CDH. They showed that the normal developmental expression of FGF18 increased during human pregnancy, rising threefold coincidentally with the onset of secondary septation; however, in the lung, contralateral to the CDH, there was reduced FGF18 expression with a failure of the normal developmental rise. There was also a failure of lung elastin to rise normally in the last 10 weeks of the pregnancy. Confirmatory findings were found in rat and sheep CDH. Tracheal occlusion (without using the cyclical occlusion/deflation model [5]) restored FGF18 mRNA and protein in the ovine model, and vitamin A restored elastin levels in the rat. These findings are not merely of interest in designing novel therapies but also add to our understanding of the basic mechanisms of lung development. Does the developing fetal lung contain cells of potential interest in stem cell research? Hennrick and associates (53) studied tracheal aspirates obtained as part of routine neonatal intensive care from intubated neonates with respiratory distress syndrome. They showed that, in more than half the patients, fibroblast-like cells could be identified that expressed mesenchymal stem cell markers and which could undergo in appropriate media adipogenic, osteogenic, and myofibroblastic differentiation. The role of these cells in the transition from acute neonatal lung injury to chronic respiratory distress requires further work. LONG-TERM CONSEQUENCES OF CHILDHOOD EVENTS
A major theme, now developing more momentum, is the importance of very early changes on long-term lung health and, by implication, the need to focus preventive efforts very early on. The Tucson Children's Respiratory Study (54) reported on lung function tracking from shortly after birth through to age 22 years in 123 participants. They showed that infants in the lowest quartile for There are a number of well-known risk factors that affect fetal lung health. Babies enter the second half of pregnancy with a normal number of airways, and in the subsequent weeks of pregnancy, adverse events may reduce airway caliber at birth. A large-population study has recently addressed this (56, 57). Maternal smoking has a direct effect on airway caliber, possibly via a reduction of alveolar tethering points (58). There are important gene–environment interactions; maternal and fetal glutathione metabolizing enzyme genotypes both interact with environmental smoke; the fetal exposure to maternal smoke is greater if the mother carries null polymorphisms, and the consequences are greater for the null child (59, 60). Maternal atopy has also been associated with impaired lung function in the newborn, although the precise mechanisms are not clear (61, 62). Maternal hypertension or preeclampsia is associated with an increased risk of transient early wheezing, persistent wheezing, and late-onset wheezing. Use of antibiotics for urinary tract infections was associated with transient early wheezing, and antibiotic administration at delivery was associated with both transient early wheezing and persistent wheezing (56). Children who had a mother with diabetes were more likely to have persistent wheezing (56). Neither amniocentesis, chorionic villus sampling, nor delivery by cesarean section was associated with the subsequent development of wheezing (56) in this study. Results from the Wheezing Illnesses Study Leidsche Rijn (WHISTLER) cohort suggested that maternal respiratory infections in pregnancy may adversely impact neonatal lung function (63). Maternal diet is another intriguing area; a large cohort study suggested that eating apples and fish in pregnancy may protect against allergic asthma and atopic disease (64), although because there are no neonatal lung function data recorded, it is not possible to determine if this is structurally mediated, or if this occurs by an effect on neonatal immune responses. Some factors appear to be additive; the risk of bronchiolitis was increased in the offspring of mothers who smoked or had asthma, and the risk to infants of asthmatic smoking mothers was 50% greater than that in the unexposed population (65). Although the arguments rage on as to whether early bronchiolitis is a cause for, or a marker of, subsequent asthma (66–69), it is likely from epidemiologic evidence (70) that bronchiolitic babies have impaired lung function before their acute bronchiolitis, and this marks them as a high-risk population for future COPD. An important genetic study has provided further evidence of the importance of early-life events. The earliest and most dramatic piece of evidence came from Barker's group, who showed a tight correlation over a large number of urban, suburban, and rural communities between death certificate data for infant mortality from respiratory causes and standardized mortality rate for COPD 50 years later (71); thus, if you came from a community where there was a high infant mortality rate from respiratory causes, there would be a high standard mortality rate for COPD five decades later. There are two important lessons from this finding: first, any understanding of COPD that ignores early-life events is likely to be flawed; and second, if you cannot measure something very accurately (and death certificate data are notoriously inaccurate), measure it many times, and you will get to the answer. The studies in very early lung function are more controversial (72, 73), but probably the transient wheezers have abnormal lung function at birth, with incomplete catch-up growth at birth, whereas the persistent wheezers had normal lung function at birth, with a decline over the first 4 years of life (73). The changes in physiology correspond to a change in airway wall histology from normal at age 1 year (74), to the established changes of asthma (eosinophilic inflammation, reticular basement membrane thickening) by age 3 (75), although it should be noted that these associations are not as fully developed as in older children and adults (76). Subsequently, it has been found in a series of overlapping cohort studies that lung function tracked from about age 3 to 4 years into middle age (77–79), and that those individuals with transient (virus-induced) wheeze had a similar and accelerated rate of lung function decline to asthmatic adults, making them a high-risk group for the development of COPD (80). One would predict that genes important in antenatal lung development would also be important in early-life lung function and COPD risk. ADAM33 (a disintegrin and metalloproteinase 33) is known to be important in lung branching morphogenesis (81), and the Manchester cohort study demonstrated that polymorphisms in this gene were related to airway resistance at ages 3 and 5 years (82). Two studies have also demonstrated that ADAM33 polymorphisms are important in COPD. The first demonstrated an association between polymorphisms that were known to convey susceptibility to COPD and rate of lung function decline in the general population (83). The same group showed that single nucleotide polymorphisms in ADAM33 also were associated with the degree of inflammation and airway reactivity in COPD (84). Thus, geneticists seeking COPD susceptibility genes would be well advised to look first in the fetal lung. Antenatal smoke exposure affects more than just airway structure. Fetal immunology is affected by maternal smoking, and by birth order, suggesting a new mechanism in which the original observation that atopic disease is less common in children with many older siblings (85) may operate. However, it should be noted that a study on more than half a million Israeli military conscripts showed that being a member of a family of four or more siblings was equally protective against asthma, regardless of position in the birth order (86), which cannot obviously be accounted for either by the effect of early infections or the immunologic effects of a previous pregnancy.
There are data showing that maternal smoking leads to lower cord blood IL-4 and IFN-
In terms of the postnatal implications, the Childhood Origins of Asthma (COAST) study showed that cord blood IFN- There has been recent work that has focused on early-life events and their long-term implications. The Prevention and Incidence of Asthma and Mite Allergy (PIAMA) birth cohort study (94) reported on the interactions between birth weight and respiratory symptoms in the first 7 years of life. As expected, the overall prevalence of respiratory symptoms, including wheeze, in the 3,628 children who remained in the study declined over time, falling from more than 20% in the first year of life, to less than 10% by age 7. In accord with other studies (95–97), this analysis found an increasing risk of wheeze in the first 5 years of life, which declined thereafter, and was no longer significant at age 7 years. The novel finding was that smoking in the infant's home increased the effect of low birth weight, such that a child with a 2.5-kg birth weight had a nearly twofold risk of symptoms compared with a child with a 4.5-kg birth weight; and in the former, smoke exposure led to a 12% increase in the prevalence of symptoms. Interestingly, there was no increase in doctor-diagnosed asthma in the low-birth-weight, smoke-exposed group. This is encouraging evidence that, at least, we are learning the following: that nonspecific respiratory symptoms, and even wheeze, in young children are not the same as asthma in atopic older children or in adults; children do not respond well to intermittent (98) or continuous inhaled corticosteroids (99–102); and they cannot be prevented from progressing to persistent (multitrigger) asthma by early treatment with inhaled corticosteroids (98–102), or indeed anything else. The benefits of giving an asthma diagnosis and thus asthma therapy to older children have been well documented. Unfortunately, in many parts of the world, the diagnosis of asthma in a young (especially younger than 2 yr) preschool child with wheeze is likely to lead to the prescription of expensive inhaled corticosteroids, which are at best of minimal value, and at worst harmful. Finally, the fact that the effects seem to die away by age 7, and there is no increased risk of symptoms, is not a cause for reassurance; as discussed above, it is exactly these children who will be in a high-risk group for COPD later in life (80). Other childhood factors may be important in later-life lung health. The traditional Mediterranean diet consists of a high intake of fruits and vegetables, predominantly whole grain bread and cereals, legumes and nuts, and olive oil as the main source of fats. In rural Crete, atopy is not uncommon, but wheeze and rhinitis are rare. Chatzi and colleagues (103) showed in a cross-sectional study that high adherence to this diet had a beneficial effect on the symptoms of asthma and rhinitis. Another group (104) confirmed a protective effect of the Mediterranean diet on chronic severe asthma in girls, and that exercise showed a dose–response protection against occasional asthma and rhinoconjunctivitis in both sexes. Interestingly, a recent study in mice has shown that exercise reduces airway inflammation and remodeling in an ovalbumin model (105). Obesity was a risk factor for chronic severe asthma in girls (104). Another group showed that eating bananas and drinking apple juice from concentrate, but not eating apples, may protect against wheezing (106). The subject of dietary manipulation in the ante- and postnatal periods has recently been reviewed (107). Allergen exposure is another important issue in long-term lung health. Lau and coworkers (108) showed that aeroallergen sensitization in the first 3 years of life was important in the persistence of wheeze. However, how to prevent sensitization is more difficult. A recent article (109) reminds us that where allergens are concerned, more may be better—a high mattress dust load was protective against inhalant allergen sensitization. Housing issues may be important: moisture damage and mold growth in the living quarters were associated with recent-onset asthma in another recent study (110). Preterm birth seems inexplicably to be becoming more, not less, common (111), and neonatologists are salvaging tiny babies with ever improving results. Furthermore, surfactant therapy and gentle ventilation strategies are reducing trauma to the newborn lung. Therefore, the effects of prematurity, independent of therapy, are becoming an increasingly important subject. The subject of chronic lung disease of prematurity has recently been reviewed (112). Previous workers have shown that preterm ventilated neonates (113, 114), and even nonventilated preterm infants, may have worsening airflow obstruction in the first year of life (115). A longitudinal study of healthy preterm infants recently reported data from the first and second years of life (116). This group had previously tested 62 preterm babies (<37 wk gestation), none of whom had received positive-pressure ventilation, or required oxygen-enriched mixtures for more than 48 hours, and compared the results with 27 term control infants (20). They used the raised volume, rapid thoracoabdominal compression technique (RVRTC), and showed that there was evidence of airway obstruction, with impairment of flow rates and expired volumes of around 30% in the preterm group. Male sex, gestational age, and weight were important predictors of lung function. The follow-up study involved 26 of these infants, and compared them with 24 term control infants. The salient findings were that, despite any particular intercurrent illnesses, there was no catch-up in forced expiratory flows during this time period, suggesting the possibility of abnormal lung development. Vital capacity remained normal throughout. Curiously, although the numbers are small and conclusions must therefore be cautious, those who had oxygen in the newborn period seemed to have accelerated lung development. Unsurprisingly, intrauterine tobacco smoke exposure had an adverse effect. The longer term effects of premature delivery on lung function have been reviewed recently (117). Again, recalling the long-term effects of small decrements in lung function early on, the good health of this cohort should not be a cause for complacency, and it should be remembered that, even if neonatalogists develop completely atraumatic means of respiratory support, the effects of preterm birth will not disappear. A recent article showing that childhood chest illness did not affect the rate of lung function decline between age 35 and 45 years (118) may appear to undermine the hypothesis that early-life events are important in long-term lung health. The authors looked at groups with childhood pneumonia, pertussis, wheeze by age 7, and wheeze onset between 8 and 16 years. They found no deviation from normal in the rate of change of lung function, measured as the difference between spirometric values at two time points 10 years apart. The cohort consisted of 18,559 babies born in a single week of 1958. Crucially, the earliest age of study was 7 years, and by the time of this study, perhaps unsurprisingly, more than 90% of participants had dropped out, leaving data on 1,156 subjects. Their findings are supported by those of other studies (78, 119), but are different from what was reported by Edwards and colleagues (80), who found a faster rate of decline than reported by others (78, 120, 121). There are three reasons why these data, although impressive, must be regarded with caution. The first is the attrition rate from their original cohort, and the possible risk of inadvertent bias. The second is that assigning the classification of childhood illness relied on parental recall at age 7, when many will have lost their symptoms (see, for example, Reference 98). In an earlier publication from this cohort (122), parents who gave a history of previous pertussis in their child had forgotten the entire episode at the time of the next questionnaire. Indeed, in that publication, the group makes exactly these two points, that sampling issues and recall bias may have affected the findings. The third issue is that a slope that depends on a line joining just two points is very vulnerable to errors in those points, caused, for example, by an undetected intercurrent illness leading to transient airflow obstruction. These errors might be expected to be randomly distributed between the two time points, and thus cancel out, but they are a source of noise in the signal. Finally, the insensitivity of spirometry to distal airway disease has been known for many years. My conclusion from this study, which was performed by a group of international high repute, is that the signal was there (early childhood respiratory disease is important), but the authors were not able to detect it. The specter of the long-term effects of childhood obesity and obstructive sleep apnea (OSA) has become increasingly important. The effects of both extend well beyond the respiratory system, but both are eminently detectable if sought. OSA is associated with neurocognitive defects and poor school performance, and potentially severe cardiovascular and metabolic complications. Pediatric OSA, and its relationship with obesity, was the subject of a recent Pulmonary Perspective (123). There are important differences compared with adult OSA: prepubertal prevalence shows no sex difference, it is more commonly associated with adenotonsillar hypertrophy than obesity, and there is a stronger (unexplained) association with respiratory allergy and asthma. The studies on the role of obesity in causing pediatric OSA are conflicting, but the balance of the evidence is surely that obesity is, at the least, a contributory factor. It has been speculated that there is a positive feedback loop between OSA leading to behavioral issues that impact on physical activity, worsening obesity. In addition, OSA may cause airway neutrophilia (124); whether this is implicated in some children whose asthma is neutrophilic remains to be determined, but perhaps polysomnography should be part of the routine diagnostic workup in these children. Childhood obesity predicts adult-onset, current asthma in women but not men in their 30s (125); unfortunately, sputum cytology was not reported in this article. There may be an impact of obesity on treatment for adenotonsillar hypertrophy. Adenotonsillectomy has a higher morbidity in obese children, and the results of surgery on physiologic disturbance are inferior (126, 127). Respiratory effects of OSA in adults were nicely summarized in a recent editorial (128). These include the effects of hyperleptinemia (causing inflammation and airway responsiveness), and the effects on smooth muscle mechanics, by the abrogation of the effects of periodic deep breathing and sighing on actin–myosin crosslinking. There is no doubt that obesity is becoming more common (129) and the prevalence is accelerating. Many children start on the "trajectory of obesity" in the very early school years, and risk factors include parental obesity, television watching and high sedentary activity time, and rapid weight gain in infancy and early and middle childhood (130, 131). Parenting styles may be important; higher paternal, but not maternal, control scores were associated with a higher body mass index in the child (132). Once adolescence is reached, adiposity is stable (133). The low levels of childhood physical activity have been documented (134). A nihilistic review suggested that there was no point in detecting obese children because there were no proven interventions that worked (135), a counsel of despair if ever there was one! However, 2007 has seen the publication of studies suggesting that an intensive, individualized intervention program (136), a parent-led, family-focused weight management program (137), simple physician advice (138), avoiding eating in large groups (139), and cognitive behavioral therapy (140) may all be effective. Surely we should be trying to detect obesity at an early stage, and evaluating programs to prevent further progression of this problem. It is clear that obesity is another factor that impacts on respiratory health, and is also undesirable in many other ways. It can and should be detected early; and the obese child, left to his/her own devices and the care of the family that has allowed obesity to develop in the first place, is most unlikely to get better spontaneously. Obesity must surely become part of any public health program that aims to ensure the well-being of future adults. RESPIRATORY INVESTIGATIONS A recent American Thoracic Society (ATS) document on pediatric respiratory research (3) highlighted the technical difficulties in researching young children, and no group, not even adolescents, is as difficult as preschool children. Infants can be sedated, and passive cooperation obtained, and a combination of threats, deceit, and bribery ensure the active cooperation of the school-age child. Previously, lung function in the preschool years has been virtually impossible to study because neither active nor passive cooperation could be obtained. However, there has been a recent explosion of new techniques, summarized in a joint ATS/European Respiratory Society statement (141). This authoritative publication, which is the last word on the subject, sets definitive standards for preschool spirometry, tidal breathing measurements, the interrupter technique, forced oscillation, multiple inert gas washout, and bronchial responsiveness. It is interesting to note that some techniques, such as spirometry, which were developed in older children and adults, have been used in preschool children previously believed to be too young to cope with the rigors of the maneuver, whereas others, such as multiple breath inert gas washout, which, although used in older children to some extent and has been a major focus of interest in preschool children, are now increasingly being used in adults. The calculation of lung clearance index (LCI) from multiple breath inert gas washout has provided us with a test that is very sensitive to distal airway dysfunction, and which, moreover, has the same normal range across the whole age spectrum. In children with cystic fibrosis (CF), LCI is more sensitive in cross-sectional studies. In the first study, LCI was compared with FEV1 in children with CF and control subjects (142) using Swedish and British data. The principal findings were that whereas there were significant negative correlations between LCI and both FEV1 and MEF25 (maximal expiratory flow), most children with a normal FEV1 had an increased (abnormal) LCI, implying that LCI was more sensitive. The normal values for LCI were similar in both countries, showing that the technique is transferable. The second article (143) compared plethysmography (specific airway resistance), spirometry (144), and LCI in a cross-sectional study. More abnormalities were detected by LCI than the other two techniques, and LCI was the only parameter that differed significantly between those who were and were not infected with Pseudomonas aeruginosa. In a longitudinal study (145), 142 children and adolescents with CF who had chronic P. aeruginosa infection were studied. Spirometry, FRC and specific airway resistance (using plethysmography), and LCI using nitrogen washout were performed. The main findings were that LCI deteriorated first, followed by FEF50 (forced expiratory flow), followed by FVC, and then FEV1. When FEV1 was normal, LCI was abnormal in 52.5%; conversely, an abnormal FEV1 was only missed by LCI in 0.5% of subjects. FEV1 stabilized at 12 years of age, whereas hyperinflation continued to worsen, and even at age 20, the slope of progression was still steepest (and thus the technique more sensitive) for LCI and flattest for FEV1. Although in older children the LCI is clearly the most sensitive test, the same is not true for infants, underscoring the need to avoid assumptions spanning the developmental stages. RVRTC and LCI were performed in 39 infants with CF and 21 control subjects (146), with a mean age of less than 1 year. Seventy-two percent of the infants with CF had an abnormality in at least one test, and, unlike the situation in older children, 15% had an abnormality on RVRTC not detected by LCI. Thus, under the age of 1 year, the two tests give complementary information, and both should be performed. Three further articles on LCI merit attention. In the first (147), LCI was measured in adults using a mass spectrometer, and a cheaper but less versatile device relying on photoacoustic spectroscopy, the latter with a very high signal-to-noise ratio but a slightly inferior response time (which will prevent its use in very young children unless the response time can be brought down [141]). The results were very comparable with the two devices, and again confirmed that a normal FEV1 does not exclude abnormalities in more sensitive tests of pulmonary function. Ultrasonic flow probes may make the measurements even more accessible by reducing the cost of the apparatus (148). Another important study addressed the question of which patients with CF should have a high-resolution computed tomography (HRCT) scan. This technique has a burden of radiation, the consequences of which are debated (149–151). However, clearly any radiation burden is undesirable without proof of improved outcome, and the challenge to demonstrate that this expensive test is worthwhile in terms of patient benefit has yet to be met. LCI and spirometry were compared with HRCT in 44 children and teenagers with CF (152). As has been shown before, spirometry was not sensitive to structural changes. A normal LCI was shown almost to exclude HRCT abnormalities, but there were a few subjects with normal HRCT and abnormal LCI. This suggests that LCI may be superior to HRCT as a monitoring tool; it is for certain safer and cheaper. However, these findings need to be confirmed by a longitudinal study, similar to those studies that confirmed the superior sensitivity of LCI over spirometry (145). Finally, sophisticated physiologic analyses, partitioning gas mixing to a conductive (Scond) and gas-exchanging (Sacin) inhomogeneity, allowed physiologic differences in CF and asthma to be highlighted, with more Sacin disturbance and less normalization of the abnormalities by bronchodilator in CF (153). We will be able to learn much more about airway physiology in the future with these sophisticated techniques. The study of normal children may not seem to be exciting, but determining how normality varies during growth and development is essential if tests are to be interpreted properly (141). Recently published articles giving new normal ranges will be important for investigators to use in future studies. In the first article, sophisticated modeling techniques were used to combine pediatric and adult datasets (3,598 subjects aged 4–80 yr) to describe the relationship between spirometry, height, and age in children and adults (154). The authors also provide ethnic corrections for these data. Smooth curves are provided, which should allow a seamless transition through puberty and into adult life, during growth and decline of lung function. Importantly, journals are likely to require all future articles to use these normal ranges, unless an adequate justification can be provided for the use of other values. The second article providing normal ranges for children addressed the six-minute-walk test. Astonishingly for such a popular test, the only previous data were for 74 normal subjects (155). Li and colleagues (156) studied 1,445 Chinese subjects, aged between 7 and 16 years. Height-specific walk-distance charts were constructed for males and females; at all ages, males walked further than females. A lower resting heart rate was predictive of a greater walk distance. The authors rightly caution about extrapolation outside their age range, and to other ethnic groups. A smaller study (n = 528) in healthy white children also provided useful normal data for the six-minute-walk distance, and showed height and age explained around 50% of the variability of distance walked (157). A study demonstrating the safety of endobronchial biopsy in children with CF (158) provoked a controversy (159–163). I will not rehash the overwhelming arguments that endobronchial biopsy is both safe and ethical during a clinically indicated fiberoptic bronchoscopy. Adequate material can be obtained (164) safely and with minimal increase in procedure time (165), and has been used to show that there are components of airway remodeling that appear to be independent of inflammation and infection (166). Taken together with recent physiologic data (167–169), it would appear that there may be a CFTR-specific defect of airway wall function, which may represent an important new therapeutic target. PEDIATRIC INTERSTITIAL LUNG DISEASE Interstitial lung disease is much rarer in children than in adults, and there are many more variants. The literature, with occasional exceptions (170, 171), consists of case reports and case series. The chILD (Children's Interstitial Lung Disease) network has published what is without doubt the most significant article in the field (172), in which a comprehensive classification, based on independent pathologic review of nearly 200 open-lung biopsies in children younger than 2 years, categorized 88% of 187 biopsies. This classification consigns the practice of "lumping" of children into "fibrosing alveolitis of children" into the obscurity from which it previously emerged. The categories that are proposed are as follows: diffuse developmental disorders (the alveolar-capillary dysplasia–acinar dysplasia spectrum), lung growth abnormalities (pulmonary hypoplasia from a variety of causes), pulmonary interstitial glycogenosis, NEHI, surfactant protein (SP) dysfunction (SpB, SpC, and ABCA3 deficiency), disorders of the normal host (mainly infectious or postinfectious, aspiration, or allergic alveolitis), disorders resulting from systemic disease processes (very few cases, mainly pulmonary hemorrhagic syndromes), disorders of the immunocompromised host (infectious and postinfectious predominant, and iatrogenic complications), and disorders masquerading as chILD (mainly pulmonary vascular and lymphatic disorders). Although the final version of this classification may be modified (173), there is no doubt that this is a major and important review. We await the corresponding article classifying disorders in 2- to 16-year-olds, so far only published in abstract form (174), and a study that explores how both classifications perform in a second population. The disappointment of the initial study is that imaging data were too sparse to be incorporated into the manuscript. In the meantime, some lessons are clear. If an open-lung biopsy is to be performed for suspected chILD, a careful protocol should be used to ensure tissue is stored for all relevant tests (175), including electron microscopy (176). Such children should have clinical data carefully and systematically recorded, and CT scans performed using standard protocols for evaluation by more than one radiologist. Biopsies from patients with chILD should be seen by more than one pathologist who has extensive experience with such disorders. These conditions are so rare that only by international collaboration, using standard protocols, are we likely to obtain large enough groups of patients for clinical trials, in particular of expensive, potentially toxic, anticytokine therapies. Furthermore, there are still more novel forms of chILD awaiting discovery; a kindred of eight children with the same father were found to have varieties of pulmonary interstitial lymphoid infiltration and autoimmune disease, including joint disease, alopecia, and a positive rheumatoid factor (177). There has also been a first report of a patient with CARD15 mutation–positive Blau syndrome who had a rather poorly characterized interstitial pneumonitis (178). The importance of SP mutations depends on the context. Although an adult study of patients with sporadic interstitial pneumonias failed to find an increase in SpC mutations (179), in 14 of 17 babies with fatal newborn respiratory distress syndrome (only two of whom were preterm) 12 had ABCA3 and 2 had SpB mutations (180). Finally, the complexity of these diseases is increased by a report suggesting that ABCA3 mutations may be modifier genes for the severity of SpC deficiency, which can present in the newborn period, or lie dormant until adult life (181). CONCLUSIONS The two most important questions for the results of any research project are "What for?" and "So what?", and it is to these two questions that this final section turns. The crucial lesson that is becoming ever clearer is that much respiratory illness in adult life is entirely predictable at or soon after birth. Bad airways in infancy lead to bad airways in adulthood and old age, and fat children become fat adults, unless interventions are put in place. A high-risk group of babies could be defined even at birth from records of birth weight and gestational age (182), maternal smoking history, maternal atopy, and maternal obstetric history. Danger signals could be picked up by an annual measurement of height, weight, and spirometry in a school-based initiative. What can be done? Environmental tobacco smoke is a recurring theme; although some progress has been made in this area (183), it is not sufficient (184). The adverse effects of indoor and outdoor environmental pollution have been discussed elsewhere (185–187). The worrying lack of exercise (other than of muscles of the thenar eminence during video games) has been highlighted. Effective public health measures should be put in place. These would include high taxes on cigarettes—would people take up smoking at $50 per cigarette?—and the banning of smoking in any public place. More controversially, I would advocate routine testing of children for tobacco smoke exposure by urine or salivary cotinine in schools or medical centers from a young age, and take a positive test as seriously as a positive finding for a substance of abuse. Why should not such a test be seen as just as routine as a blood pressure check? Obesity could be tackled by removing advertisements glorifying the serving of huge portions of fatty foods, and listing of all ingredients and their calorific values made mandatory. Obesity can be easily detected in schools, including at first entry to the school system; tackling it may be harder, although as documented above, a number of possible strategies may be useful. At the very least, extra exercise sessions should be mandatory for obese children, if only to impress on all concerned that obesity is a serious problem. Should childhood obesity in the child with no underlying syndrome be considered to be a form of child abuse if it is persistent, and no discernable effort is made to tackle it? A package of measures described in this review could have the potential to improve lung health over the years, and halt the COPD epidemic. However, these measures have one thing in common: they all involve short-term political pain, for long-term gain, and since politicians the world over rarely think beyond the next election, such unpopularity is rarely likely to be risked. It is much easier to set targets for others than introduce unpopular measures oneself, and easiest of all to generate reports that are "a triumph for vapid political style and flabby prose over meaningful substance" (from my choice for must-read article of the year) (188). FOOTNOTES Conflict of Interest Statement: A.B. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. Received in original form January 7, 2008; accepted in final form January 7, 2008 REFERENCES
|
maxFRC shortly after birth had the lowest FEV1/FVC ratio at age 22 years, after adjustment for height, weight, age, and sex. They also make the point that prevention of chronic obstructive pulmonary disease (COPD) may have to start at birth or even antenatally. A longitudinal birth cohort study in more than 5,000 adults has shown that, at age 31 years, each 500-g increment of birth weight resulted in a 53.1-ml increase in FEV1 and a 52.5-ml increment in FVC (95% confidence intervals: 38.4–67.7 and 35.5–69.4, respectively) (
(
production. TLR3- and TLR4-mediated signaling of TNF-