Lung Development and Susceptibility to Ventilator-induced Lung Injury
http://www.100md.com
美国呼吸和危急护理医学 2005年第4期
Lung Biology Program, Departments of Critical Care Medicine and Pediatrics, Hospital for Sick Children
Departments of Anesthesia, Pediatrics, Physiology
the Interdepartmental Division of Critical Care Medicine, University of Toronto, Toronto
Intestinal Disease Research Program, Department of Medicine, McMaster University, Hamilton
Department of Pediatrics (Critical Care Unit, Children's Hospital of Western Ontario), University of Western Ontario, London, Ontario, Canada
ABSTRACT
Rationale: Ventilator-induced lung injury has been predominantly studied in adults. Objectives: To explore the effects of age and lung development on susceptibility to such injury. Methods: Ex vivo isolated nonperfused rat lungs (infant, juvenile, and adult) were mechanically ventilated where VT was based on milliliters per kilogram of body weight or as a percentage of the measured total lung capacity (TLC). In vivo anesthetized rats (infant, adult) were mechanically ventilated with pressure-limited VTs. Allocation to ventilation strategy was randomized. Measurements: Ex vivo injury was assessed by pressureeCvolume analysis, reduction in TLC, and histology, and in vivo injury by lung compliance, cytokine production, and wet- to dry-weight ratio. Main Results: Ex vivo ventilation (VT 30 ml · kgeC1) resulted in a significant reduction (36.0 ± 10.1%, p < 0.05) in TLC in adult but not in infant lungs. Ex vivo ventilation (VT 50% TLC) resulted in a significant reduction in TLC in both adult (27.8 ± 2.8%) and infant (10.6 ± 7.0%) lungs, but more so in the adult lungs (p < 0.05); these changes were paralleled by histology and pressureeCvolume characteristics. After high stretch in vivo ventilation, adult but not infant rats developed lung injury (total lung compliance, wet/dry ratio, tumor necrosis factor ). Surface video microscopy demonstrated greater heterogeneity of alveolar distension in ex vivo adult versus infant lungs. Conclusion: These data provide ex vivo and in vivo evidence that comparable ventilator settings are significantly more injurious in the adult than infant rat lung, probably reflecting differences in intrinsic susceptibility or inflation pattern.
Key Words: infant lung injury mechanical ventilation pediatric ventilator-induced lung injury
Lung injury, or worsened lung injury during mechanical ventilation, is clearly a matter of major concern. Clinicians caring for ventilated patients, as well as researchers investigating the responsible mechanisms, attempt to understand the optimal approach to ventilation to lessen lung injury. A current approach whereby excessive tidal volumes (VTs) are avoided (1, 2) has evolved over time on the basis of laboratory studies performed largely in adult animals (3). Furthermore, the demonstration that the manner of application of mechanical ventilation directly affects survival attests to the importance of this area of research and has been validated in a landmark randomized clinical trial in adults (4).
Compared with adults, pediatric patients demonstrate a spectrum of lung development spanning neonatal, infant, youth, and adult stages. Beyond neonatal age, there are important differences between infant and adult lungs (e.g., alveolar structure, matrix composition, angiogenesis) (5). In fact, maturation in the human lung continues well after the newborn period until between the ages of 2 and 8 years (5). Although lung injury in preterm neonates has been extensively investigated and characterized (6), there has been limited laboratory investigation (7eC9) and no prospective clinical investigation of ventilator-associated lung injury affecting infancy or youth. Instead, clinical research concerning mechanical ventilation in infants and young children (nonneonatal, nonadult) has focused not on VT or airway pressure but on adjuvant therapies, including inhaled nitric oxide (10), surfactant (11), and the use of high-frequency oscillation (12). As a result, evidence-based age-specific guidelines for the use of conventional mechanical ventilation in pediatric patients have not been possible; furthermore, any recommendations that exist have been extrapolated from adult data (13, 14).
An empiric sense that infant lungs might be more susceptible to ventilator-induced lung injury (VILI) compared with adult lungs is supported by a single laboratory investigation that compared juvenile with adult rabbits (7). That report, which demonstrated greater injury in the lungs of younger animals, concluded that the findings may have reflected the larger VTs delivered because of greater respiratory system compliance (7). However, there is an important reason to believe that infant lungs may be less susceptible to VILI. VT is usually expressed in terms of milliliters per kilogram of body weight, but the ratio of lung volume compared with body weight varies with development. In rats, it increases (15), and in humans, it may decrease (16). Rather than attempt to define locally applied mechanical stress, the current study evaluated the pulmonary responses to comparable levels of VT on the basis of body weight (ml · kgeC1) and lung volume (ml/ml, % total lung capacity [TLC]), as well as on inflation pressure.
Our main objective in this study was to explore whether lung maturation beyond the neonatal age range has an effect on susceptibility to VILI in the rat lung. The rat was chosen because of the well-characterized developmental morphology (15, 17, 18), which, after 4 days of age, is comparable to human lung development (19, 20). To remove any influence of age-related chest wall compliance or lungeCheart interaction, an ex vivo, isolated, nonperfused rat lung model was studied (21, 22). We then reproduced our experiments in the in vivo anesthetized rat. Some of the results of these studies have been previously reported in the form of an abstract (23).
METHODS
Male Sprague-Dawley rats of three maturity levels: adult (250eC360 g, age 75 days), juvenile (72eC100 g, age 35 days), and infant (26eC45 g, age 17 days) were used. The study was conducted according to the guidelines of the Canadian Council for Animal Care and was approved by the Animal Care Committee of the Hospital for Sick Children. Complete details of the experimental protocol are provided on the online supplement. The animals were anesthetized and ventilated (24) with peak inspiratory pressure of 9 to 10 cm H2O, positive end-expiratory pressure of 1 cm H2O, inspired oxygen of 0.21, and respiratory rates of 45 (adult), 60 (young), and 80 (infant) breaths/minute. For ex vivo experiments, lungs and heart were removed en bloc and suspended in a warm, humidified chamber (21, 25, 26).
Experimental Outline
After determination of baseline TLC and pressureeCvolume (P-V) curves, lungs in Series 1 and 2 were randomly allocated to receive either no ventilation (control) or mechanical ventilation for 60 minutes. In Series 1, ex vivo lungs were ventilated with a VT of 30 ml · kgeC1, a positive end-expiratory pressure of 0, a respiratory rate of 40 minuteseC1, an FiO2 of 0.21, and an FiCO2 of 0.05 for 60 minutes. In Series 2, ex vivo lungs were ventilated with a VT set to 50% of baseline TLC. After the experiments, lung weight was recorded. In Series 3, anesthesia was maintained and the intact in vivo infant and adult animals ventilated with a peak inspiratory pressure of 20 or 30 cm H2O for 90 minutes, and compared with nonventilated control animals. In the in vivo experiments, respiratory rate was the same in all groups (34 minuteseC1), and added dead space was used to obtain comparable PaCO2 levels (35eC45 mm Hg) in each group. In Series 4, lungs were ventilated as in Series 2, and with a respiratory rate of 30 minuteseC1. Alveolar inflation and deflation were measured using inverted microscopy and analyzed offline. Static P-V curves were constructed, before and after ventilation (21, 27). Lungs were excluded from the experiments in the event of air leaks or technical problems with preparation. After in vivo experiments (Series 3), the decrement in quasi-static respiratory compliance, bronchoalveolar lavage concentration of tumor necrosis factor (TNF-), and the lung wet/dry weight were determined.
Morphologic Analysis
The left lung was isolated and inflation-fixed in formalin (inflation pressure 20 cm H2O for 48 hours) (22). An assessment was performed independently by two observers, who were blinded as to group allocation and the mean values reported (22).
Statistical Analysis
The data are expressed as mean ± SD. Statistical comparisons were performed with either analysis of variance followed by Student-Neumann-Keuls tests (Series 1), or with Student's t test (Series 2, 3). TNF- (Series 3) was analyzed using the nonparametric Kruskal-Wallis test, and post hoc analysis was with Dunn's method. Histology scores between observers were compared using interclass correlation coefficient (0.88, 95% confidence intervals, 0.817eC0.924). We considered differences significant where p was less than 0.05.
RESULTS
Experimental Series 1: Ex Vivo VT Determined by Body Weight
Baseline characteristics.
A total of 54 animals were selected; 50 were randomized, and 48 completed the protocol. Baseline variables are reported in Table 1. The TLC corrected for body weight was ranked as follows: infant > juvenile > adult (Table 1; p < 0.05). This pattern was conserved for baseline chord compliance, which was also corrected for body weight (Table 1; p < 0.05). The baseline peak inspiratory pressure, recorded 10 minutes after commencement of ventilation, was significantly higher in the adult group than in the other groups, with a rank order as follows: adult > juvenile infant lungs (Table 1; p < 0.05).
Impact of ventilation on lung mechanics.
After 60 minutes of ventilation with high VT, the TLC decreased significantly from the baseline values in the adult and juvenile groups, but not in the infant group (Figure 1). The TLC (expressed in ml · kgeC1) decreased from 39.2 ± 3.0 to 25.0 ± 3.6 ml · kgeC1, a reduction of 36.0 ± 10.1%, in the adult group (Figure 1; p < 0.05). In the juvenile group, the TLC decreased from 56.4 ± 4.7 to 47.5 ± 6.1 ml · kgeC1, a reduction of 15.4 ± 10.1% (Figure 1; p < 0.05). In the infant group, the TLC was not significantly reduced after ventilation (Figure 1; 79.8 ± 9.2 vs. 75.9 ± 7.4 ml · kgeC1, p = 0.32). The rank order of high VTeCinduced decrease in TLC was as follows: adult > juvenile infant (Figure 1; p < 0.05).
In addition to the TLC data, the complete P-V characteristics are presented for each group. Because the P-V characteristics change with age (28), the baseline and final P-V curves for each of the groups were expressed in the following three ways: pressure (cm H2O) versus uncorrected volume (ml; Figure 2A), pressure (cm H2O) versus volume per kilogram of animal body weight (ml · kgeC1; Figure 2B) (7), and as pressure (cm H2O) versus volume expressed as a percentage of the preventilation TLC (Figure 2C) (29). Comparison of these P-V curves demonstrates the following features, which are independent of how the inflation and deflation volumes are expressed. First, the P-V curves displayed postventilation downward shift, reflecting a reduction in chord compliance as follows: adult > juvenile > infant (Figure 2). Second, there was no inflection point demonstrable in any of the preventilation curves, and after ventilation, a lower inflection point developed in the adult lungs, but not in the infant or juvenile lungs (Figure 2).
Impact of ventilation on lung histology.
There was no difference in alveolar injury, expressed as hyaline membrane score, among the nonventilated (control) adult, juvenile, or infant lungs (Figure 3). The hyaline membrane score was greater in ventilated than in nonventilated lungs in the adult group only, and not in the juvenile or infant groups. In ventilated lungs, the rank order of hyaline membrane score was as follows: adult > juvenile infant groups (Figure 3; p < 0.05). There were no differences detected in hyaline membrane score between proximal and distal airspaces in any of the groups (data not presented). Airway injury scores were not different among the nonventilated (control) lungs from adult juvenile or infant animals (Figure 4). The airway injury score was significantly greater in ventilated lungs compared with nonventilated lungs in adult, but not in juvenile or infant groups (Figure 4; p < 0.05). In ventilated lungs, the rank order of airway injury score was as follows: adult > juvenile infant groups (Figure 4; p < 0.05). There were no differences detected in airway injury score between proximal and distal airways (data not presented).
Series 2: Ex Vivo VT Determined by TLC
In Series 2, VT was dictated not by body weight (i.e., ml · kgeC1) but by the baseline lung volume (i.e., VT administered was 50% of baseline TLC). Pilot studies indicated that this approach was associated with lower peak airway pressures in the adult versus infant lungs. The greatest differences in Series 1 were between the adult and the infant lungs, and so these groups were the focus of Series 2.
Baseline characteristics.
In this series, seven adult and seven infant animals were selected; six from each group were studied, and all completed the experiments. The baseline variables are reported in Table 2.
Impact of ventilation on lung mechanics.
Despite the higher VT (ml · kgeC1) and peak inspiratory pressure in the infant lungs, 60 minutes of mechanical ventilation with a VT at 50% of baseline TLC was associated with a significantly smaller reduction in TLC in the infant versus the adult lungs (Figure 5; p < 0.05). In the adult lungs, TLC corrected for body weight decreased from 39.8 ± 2.88 to 28.5 ± 2.82 ml · kgeC1 (Figure 5; p < 0.05), whereas in the infant group, the TLC decreased from 68.3 ± 10.5 to 61.1 ± 10.9 ml · kgeC1 (Figure 5; p < 0.05). The percentage of reduction in TLC (% baseline value) was greater in the adult versus the infant lungs (28.7 ± 2.84 vs. 10.6 ± 7.0%; Figure 5; p < 0.05). This pattern was maintained with expression of the P-V curves in terms of the following: pressure (cm H2O) versus uncorrected volume (ml), pressure (cm H2O) versus volume corrected for body weight (ml · kgeC1), and as pressure (cm H2O) versus volume expressed as a percentage of the baseline volume at the same static pressure (Figure 6).
Impact of ventilation on lung histology.
Hyaline membrane scores and airway injury scores were both significantly greater after ventilation in the adult versus the infant lungs (Figure 7; see Figure E1 in the online supplement). No significant differences in hyaline membrane scores or airway injury score were observed between proximal and distal areas (data not presented).
Series 3: In Vivo Ventilation
Baseline characteristics.
Thirty-two animals were selected, and all completed the experiment. The baseline variables are reported in Table 3. When ventilated with 20 cm H2O, the resultant VTs in the infants versus adults were 37.5 ± 1.2 versus 18.4 ± 1.3 ml · kgeC1, respectively (Table 3; p < 0.05). Similarly, when ventilated with 30 cm H2O, the resultant VTs in the infants versus adults were 42.4 ± 4.4 compared with 29.7 ± 3.8 ml · kgeC1 (Table 3; p < 0.05).
Impact on lung mechanics and edema.
After 90 minutes of mechanical ventilation, no change in total lung compliance was detected in the adult animals ventilated with a pressure of 20 cm H2O, but a significant reduction in lung compliance occurred in those ventilated with a pressure of 30 cm H2O (Figure 8). Despite being ventilated with comparatively higher VTs (in terms of ml · kgeC1), no reductions in lung compliance occurred in any of the infant animals, whether ventilated with 20 or 30 cm H2O (Figure 8). The magnitude of the effect of ventilation on compliance in the adult and infant animals paralleled the changes in lung wet/dry ratio (Figure 9).
Impact on pulmonary TNF-.
The volume of bronchoalveolar lavage fluid retrieved was similar within each age group. An illustrative proinflammatory cytokine, TNF- (30), was not detected in nonventilated control animals and was significantly greater in ventilated adults ventilated with 30 cm H2O versus all other groups, with no other among-group differences (Figure 10; p < 0.05).
Series 4: Ex Vivo Ventilation—Lung Imaging
Three animals from each group (infant 43 ± 1.8 g, adult 331 ± 33 g) completed the experiment. The mean TLC (infant 38.2 ± 3.2 ml · kgeC1, adult 69 ± 7.2 ml · kgeC1) was similar to the respective values in Series 1 and 2. Mean alveolar diameter at the end-expiration was 39.7 ± 8.4 e in adults compared with a mean of 35.3 ± 4.5 e in infants (p = not significant). Mean alveolar diameter at end-inspiration was significantly greater in the adult versus infant lungs (77.3 ± 6.7 vs. 57.3 ± 2.0 e, p < 0.05). The mean variance of the alveolar diameter at the end of inspiration was significantly greater in adult versus infant lungs (824 ± 240 vs. 318 ± 40, p < 0.05; Figure 11).
DISCUSSION
The principal finding of this study is that, whether VT is based on body weight (ml · kgeC1) or baseline TLC (%), comparable ventilator settings resulted in greater injury in adult versus infant (nonneonatal, nonadult) lungs. The lessened lung injury in infants was confirmed in vivo, despite exposure to larger VT (relative to body weight) and higher peak inspiratory pressure. In addition, adult lungs appeared to have a more heterogeneous distribution of alveolar ventilation. Although the current study did not directly address the issues of local lung stress or strain (31), the overall data may suggest that intrinsic properties of the lung are responsible for the lesser susceptibility in infants.
Development and Susceptibility to Lung Injury
Laboratory or clinical studies of VILI have been performed mostly in surfactant-deficient preterm newborns (32, 33) or in adults (3, 4, 22, 34). Two articles have recently reported important findings that may relate to mechanisms of VILI in nonpremature newborn models (35, 36). Newborn rats exposed to high VT demonstrate less cytokine mRNA production compared with adult animals (36). In addition, Rau and coworkers (35) have demonstrated that, in a porcine model, the newborn has different surfactant composition—and superior dynamic function—than the adolescent.
Such comparative studies are important because they recognize the potential impact of development on susceptibility to VILI between newborn versus adult lungs. However, beyond the newborn or neonatal period, only three laboratory studies have addressed VILI in young animals (7eC9). The only study (7) that directly compared juveniles with adults demonstrated that, with comparably elevated peak inspiratory pressures, younger animals developed more lung injury. The authors speculated that, because of greater respiratory system compliance, the younger animals might have been exposed to disproportionately larger VT (7).
The current study focuses on the infant versus adult susceptibility, where infancy in this model is characterized by almost complete alveolarization (i.e., having attained the adult number of alveoli) (15, 18), and with the principal differences between infant (nonneonate) and adult lungs being the degree of matrix development (28, 37).
Choice of VT
Where VT is considered in the clinical context, it is usually specified as milliliters per kilogram of body weight. Although the relationship between VT, body weight, and ventilator-associated lung injury is complex in adults (38), there is an additional layer of complexity in children. This is because even where weights can be calculated with high degrees of accuracy, the ratio of lung volume (TLC) to body weight changes with age and development. The ratio of TLC to body weight in the present study decreased over maturation from 74.0 ± 8.8 ml · kgeC1 in infant, to 53.4 ± 5.6 ml · kgeC1 in juvenile, and to 39.7 ± 4.0 ml · kgeC1 in adult groups (data averaged from Tables 1 and 2). These findings are comparable to previously reported data for the developing rat (15, 28, 39, 40). Although the optimal measurement of lung volume in young children is not established, the available data suggest that TLC/body weight ratio increases with age from 52 ± 13 ml · kgeC1 in infants to 87 ± 11 ml · kgeC1 in older children (16, 29).
Potential Mechanisms
There are several possible explanations for the current findings. The absolute VT was different among the groups. However, in Series 1, the VT was identical in terms of milliliters per kilogram of body weight and in Series 2 was dictated by TLC (50% TLC). In addition, airway pressure, although slighter higher in the adult group in Series 1, was significantly lower ( 2 cm H2O) in the adult group in Series 2. Thus, neither differences in VT nor in airway pressure explain the among-group differences in injury; in fact, the study was biased against finding worse injury in the adult lungs in Series 2, because both the VT (in ml · kgeC1) and airway pressures were lower in the adult lungs.
The inclusion of in vivo experiments provides additional important information, because the presence of the intact chest wall, pulmonary perfusion with intact blood, as well as the potential for cardiopulmonary interactions corroborate the ex vivo findings. There are significant differences in baseline TLC, which, when corrected for body weight, are far larger in infant versus adult lungs (Tables 1 and 2). This result could certainly explain the findings in Series 1 where VT based on body weight would occupy a smaller fraction of TLC in infants versus adults and therefore cause less stretch-induced injury, but this could not explain the differences observed in Series 2. Lung compliance represents the interaction of pressure and volume and determines how a given VT alters pressure and vice versa. However, comparing baseline compliance among different age groups is complex, and although correction for either body weight or lung weight has been reported (41), there is no universally agreed-on approach. Both approaches have been used in the current experiments, and when corrected for body weight, the compliance is greater in infants, but when corrected for lung weight, is greater in adults. Overall, the complexity of correcting or normalizing lung parameters for either body weight or lung weight is apparent from the current study, as well as from previous publications (15, 28, 39eC41). It seems logical therefore to expect a different impact of "body weighteCdirected" versus "lung volumeeCdirected" prescription of VT.
The pattern of alveolar distension might provide additional explanation. Real-time direct visualization of the lung has been recently described (42), and has provided important insights into alveolar micromechanics. Notwithstanding the sampling limitations of such surface visualization of subpleural alveoli (43), the current study describes a greater degree of heterogeneous subpleural alveolar inflation in adult versus infant lungs. This finding raises, but does not prove, the possibility that either underdistension (22) and/or overdistension (44) may account for the differences in injury between the two groups. Furthermore, the downward and rightward shift of the adult P-V curves after ventilation, as well as the appearance of a lower inflection point, may support the previously described findings and suggest an atelectasis-associated mechanism of injury in the adult lung (Figures 2 and 6) (45).
It is possible that the infant lung, which at 17 days has almost full alveolarization but with smaller alveoli, has an increased quantity of surfactant relative to the alveolar size. If true, this could play a role in the increased uniformity of inflation and lesser susceptibility to VILI observed in the infant lungs. Unfortunately, surfactant analysis was not performed in the current study.
An integrated explanation of these phenomena may involve appreciation of the differences in the matrix of infant versus adult lungs. The levels of collagen and elastin, the major matrix components, undergo significant changes between 4 and 40 days (28). Collagen concentration increases linearly over the interval from infancy to adulthood, whereas elastin concentration increases 10-fold over the first 20 days and rises less rapidly thereafter (46). In addition, the cross-linking of collagen changes with age. Furthermore, data from lung slice experiments indicate that mechanical stretch changes lung matrix properties (47). Thus, it is possible that the age-related change in ratio of elastin to collagen could contribute to altered susceptibility to stretch-induced lung injury. Finally, the recent studies cited previously (35, 36) may provide additional explanation. If the relatively blunted cytokine response (36) or enhanced surfactant function (35) that were reported in the newborn are also present in older infants, then the lowered susceptibility to ventilator-associated lung injury of infants may be, in part, attributable to these phenomena.
Pulmonary Cytokine Release
The role of cytokines in general and TNF- in particular in the genesis of VILI is controversial (48, 49). The observation that the concentration of TNF- increases significantly after high VT ventilation in vivo in the adult but not the infant (Figure 10) mirrors a previous report (36) where mechanical ventilation promoted less production of cytokine in newborn compared with adult rats. That finding may be explained by a difference in immunologic maturation between adult versus infant rats, or by the known differences in the matrix composition that may affect the way that global lung stretch is transmitted to individual cells.
Choice of Experimental Models and Study Limitations
The use of an ex vivo preparation alone would limit the applicability of the data. The current data, however, include in vivo confirmation, and the ex vivo data control for the impact of pulmonary vascular pressure or flow on injury (50, 51). The TLC in this study was measured in nondegassed lung. In fact, we measured the volume above the FRC volume. Furthermore, the value of FRC may differ across progressive stages of lung development, and an increased ratio of FRC to TLC has been described in newborn versus adult rats (41). We know of no data regarding true FRC values in the isolated nonperfused rat lung at infancy, but (at any age) they are almost certainly far smaller than in vivo values. Because each lung served as its own control, baseline differences in FRC among the groups should not alter the conclusions.
All animal models of human development are necessarily limited. Although there are important differences in development between the human and rat lung (52), the rat lung provides a good model for human lung development after the age of 4 days (19). At 1 month, the human lung is morphologically comparable to a rat lung aged 1 week, and at 2 to 5 years, the human lung is comparable to a rat lung aged 17 to 21 days (20). It is possible that development confers differences in viability on ex vivo lung preparations. Furthermore, data indicate that infants may be less susceptible than adults to organ ischemia or reperfusion (53). However, the current ex vivo experiments were performed within the known limits of lung viability (54), and such issues would not affect the in vivo data.
We have tested three different sets of comparisons (i.e., VT according to body weight, TLC, or by pressure); however, the current study did not directly address the concepts of stress or strain at the local tissue level. The central nature of these concepts has been commented on previously (31). Lung weight was measured in Series 2, and the higher ratio of VT (ml) to lung weight (g) in the adult versus the infant lungs raises the possibility that stress per unit of lung tissue mass might contribute to the increased injury in the adult lungs (Table 2).
Because neither airway resistance nor plateau pressure were recorded, it is possible that the lesser injury observed in the infant may be attributed to higher airway resistance or to auto-positive end-exipirator pressure. However, as previously suggested by Gomes and coworkers (40), airway resistance in rats, when corrected for weight, is not significantly different between the infant and adult. In addition, the length of the endotracheal tube was designed to abolish any differences in endotracheal tube resistance (see online supplement for details). Furthermore, from the clinical perspective, lung mechanical stretch is extrapolated from the magnitude of VT and/or plateau pressure. Thus, inclusion of plateau pressure, especially in the ex vivo preparations, should be considered in future work.
Potential Significance
This study, the first study to make age-specific comparisons of VILI between infants and adults, suggests four potential lessons. First, infant lungs may be less susceptible to VILI when exposed to comparable ventilation settings of VT or pressure. Second, development or propagation of stretch may be different in infant than adult lungs. Furthermore, if the differences observed in inflation pattern (Figure 11) are confirmed in other studies, then the utility of positive end-expiratory pressure or recruitment maneuvers may be different in infants versus adults. Third, we suggest caution in extrapolating recommendations about ventilator settings from adult studies to children, especially where the VT is prescribed on the basis of milliliters per kilogram of body weight. Finally, because understanding preclinical phenomena is an important factor in the design of successful clinical trials (55), studies such as the current report might eventually influence clinical evaluation of pediatric ventilatory strategies that are distinct from adult studies.
Acknowledgments
The authors thank Dr. A.C. Bryan for his insightful comments and Derek Stephens (Program in Population Health Sciences, Hospital for Sick Children) for statistical expertise.
This article has an online supplement, which is accessible from this issue's table of contents at www.atsjournals.org
REFERENCES
Marini JJ, Gattinoni L. Ventilatory management of acute respiratory distress syndrome: a consensus of two. Crit Care Med 2004;32:250eC255.
Tobin MJ. Advances in mechanical ventilation. N Engl J Med 2001;344:1986eC1996.
Dreyfuss D, Saumon G. Ventilator-induced lung injury: lessons from experimental studies. Am J Respir Crit Care Med 1998;157:294eC323.
Acute Respiratory Distress Syndrome Network. Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung injury and the acute respiratory distress syndrome. N Engl J Med 2000;342:1301eC1308.
Chernick V, Boat TF, Kendig EL, editors. Kending's disorders of the respiratory tract in children, 6th ed. Philadelphia: W.B. Saunders; 1998.
Speer CP. Inflammatory mechanisms in neonatal chronic lung disease. Eur J Pediatr 1999;158:S18eCS22.
Adkins WK, Hernandez LA, Coker PJ, Buchanan B, Parker JC. Age effects susceptibility to pulmonary barotrauma in rabbits. Crit Care Med 1991;19:390eC393.
Peevy KJ, Hernandez LA, Moise AA, Parker JC. Barotrauma and microvascular injury in lungs of nonadult rabbits: effect of ventilation pattern. Crit Care Med 1990;18:634eC637.
Coker PJ, Hernandez LA, Peevy KJ, Adkins K, Parker JC. Increased sensitivity to mechanical ventilation after surfactant inactivation in young rabbit lungs. Crit Care Med 1992;20:635eC640.
Okamoto K, Hamaguchi M, Kukita I, Kikuta K, Sato T. Efficacy of inhaled nitric oxide in children with ARDS. Chest 1998;114:827eC833.
Dekowski SA, Holtzman RB. Surfactant replacement therapy: an update on applications. Pediatr Clin North Am 1998;45:549eC572.
Hoehn T, Krause M, Krueger M, Hentschel R. Treatment of respiratory failure with inhaled nitric oxide and high-frequency ventilation in an infant with respiratory syncytial virus pneumonia and bronchopulmonary dysplasia. Respiration (Herrlisheim) 1998;65:477eC480.
Matthews BD, Noviski N. Management of oxygenation in pediatric acute hypoxemic respiratory failure. Pediatr Pulmonol 2001;32:459eC470.
Hirschl RB. Support of respiratory failure in the pediatric surgical patient. Curr Opin Pediatr 2002;14:459eC469.
Burri PH, Dbaly J, Weibel ER. The postnatal growth of the rat lung: I. Morphometry. Anat Rec 1974;178:711eC730.
Thorsteinsson A, Larsson A, Jonmarker C, Werner O. Pressure-volume relations of the respiratory system in healthy children. Am J Respir Crit Care Med 1994;150:421eC430.
Burri PH. Postnatal growth and maturation of the lung. Chest 1975;67:2SeC3S.
Burri PH. The postnatal growth of the rat lung: 3. Morphology. Anat Rec 1974;180:77eC98.
Cherukupalli K, Larson JE, Puterman M, Sekhon HS, Thurlbeck WM. Comparative biochemistry of gestational and postnatal lung growth and development in the rat and human. Pediatr Pulmonol 1997;24:12eC21.
Zeltner TB, Burri PH. The postnatal development and growth of the human lung: II. Morphology. Respir Physiol 1987;67:269eC282.
Tremblay L, Valenza F, Ribeiro SP, Li J, Slutsky AS. Injurious ventilatory strategies increase cytokines and c-fos m-RNA expression in an isolated rat lung model. J Clin Invest 1997;99:944eC952.
Muscedere JG, Mullen JB, Gan K, Slutsky AS. Tidal ventilation at low airway pressures can augment lung injury. Am J Respir Crit Care Med 1994;149:1327eC1334.
Kornecki A, Tsuchida S, Engelberts D, Frndova H, Belik J, Kavanagh BP. Ventilator-associated lung injury: are infants as susceptible as adults Am J Respir Crit Care Med 2004;169:A787.
Duggan M, McCaul CL, McNamara PJ, Engelberts D, Ackerley C, Kavanagh BP. Atelectasis causes vascular leak and lethal right ventricular failure in uninjured rat lungs. Am J Respir Crit Care Med 2003;167:1633eC1640.
Veldhuizen RA, Tremblay LN, Govindarajan A, van Rozendaal BA, Haagsman HP, Slutsky AS. Pulmonary surfactant is altered during mechanical ventilation of isolated rat lung. Crit Care Med 2000;28:2545eC2551.
Ribeiro SP, Rhee K, Tremblay L, Veldhuizen R, Lewis JF, Slutsky AS. Heat stress attenuates ventilator-induced lung dysfunction in an ex vivo rat lung model. Am J Respir Crit Care Med 2001;163:1451eC1456.
Venegas JG, Harris RS, Simon BA. A comprehensive equation for the pulmonary pressure-volume curve. J Appl Physiol 1998;84:389eC395.
Nardell EA, Brody JS. Determinants of mechanical properties of rat lung during postnatal development. J Appl Physiol 1982;53:140eC148.
Fagan DG. Post-mortem studies of the semistatic volume-pressure characteristics of infants' lungs. Thorax 1976;31:534eC543.
Chiumello D, Pristine G, Slutsky AS. Mechanical ventilation affects local and systemic cytokines in an animal model of acute respiratory distress syndrome. Am J Respir Crit Care Med 1999;160:109eC116.
Marini JJ. Relative importance of stretch and shear in ventilator-induced lung injury. Crit Care Med 2004;32:302eC304.
Clark RH, Gerstmann DR, Jobe AH, Moffitt ST, Slutsky AS, Yoder BA. Lung injury in neonates: causes, strategies for prevention, and long-term consequences. J Pediatr 2001;139:478eC486.
Bhuta T, Henderson-Smart DJ. Elective high-frequency oscillatory ventilation versus conventional ventilation in preterm infants with pulmonary dysfunction: systematic review and meta-analyses. Pediatrics 1997;100:E6.
Dreyfuss D, Soler P, Basset G, Saumon G. High inflation pressure pulmonary edema: respective effects of high airway pressure, high tidal volume, and positive end-expiratory pressure. Am Rev Respir Dis 1988;137:1159eC1164.
Rau GA, Vieten G, Haitsma JJ, Freihorst J, Poets CF, Ure BM, Bernhard W. Surfactant in newborn compared to adolescent pigs: adaptation to neonatal respiration. Am J Respir Cell Mol Biol 2003;30:694eC701.
Copland IB, Martinez F, Kavanagh BP, Engelberts D, McKerlie C, Belik J, Post M. High tidal volume ventilation causes different inflammatory responses in newborn versus adult lung. Am J Respir Crit Care Med 2004;169:739eC748.
Tanaka R, Ludwig MS. Changes in viscoelastic properties of rat lung parenchymal strips with maturation. J Appl Physiol 1999;87:2081eC2089.
Thompson TB, Hayden D, Matthay MA, Brower R, Parsons PE. Clinicians' approaches to mechanical ventilation in acute lung injury and ARDS. Chest 2001;120:1622eC1627.
Bartlett D Jr, Areson JG. Quantitative lung morphology in newborn mammals. Respir Physiol 1977;29:193eC200.
Gomes RF, Shardonofsky F, Eidelman DH, Bates JH. Respiratory mechanics and lung development in the rat from early age to adulthood. J Appl Physiol 2001;90:1631eC1638.
Fisher JT, Mortola JP. Statics of the respiratory system in newborn mammals. Respir Physiol 1980;41:155eC172.
Steinberg J, Schiller HJ, Halter JM, Gatto LA, Dasilva M, Amato M, McCann UG, Nieman GF. Tidal volume increases do not affect alveolar mechanics in normal lung but cause alveolar overdistension and exacerbate alveolar instability after surfactant deactivation. Crit Care Med 2002;30:2675eC2683.
Kavanagh BP. Lung recruitment in real time: learning was never so easy. Am J Respir Crit Care Med 2003;167:1585eC1586.
Dreyfuss D, Saumon G. Barotrauma is volutrauma, but which volume is the one responsible Intensive Care Med 1992;18:139eC141.
Martin-Lefevre L, Ricard JD, Roupie E, Dreyfuss D, Saumon G. Significance of the changes in the respiratory system pressure-volume curve during acute lung injury in rats. Am J Respir Crit Care Med 2001;164:627eC632.
Brody JS, Therlbeck WM. Development, growth, and aging of the lung. In: Fishman MP, editor. Handbook of physiology: the respiratory system—mechanics of breathing. Vol. 3. Bethesda, MD: American Physiological Society; 1986. pp. 355eC386.
Al-Jamal R, Ludwig MS. Changes in proteoglycans and lung tissue mechanics during excessive mechanical ventilation in rats. Am J Physiol Lung Cell Mol Physiol 2001;281:L1078eCL1087.
Dreyfuss D, Ricard JD, Saumon G. On the physiologic and clinical relevance of lung-borne cytokines during ventilator-induced lung injury. Am J Respir Crit Care Med 2003;167:1467eC1471.
Wilson MR, Choudhury S, Takata M. Pulmonary inflammation induced by high stretch ventilation is mediated by tumor necrosis factor signalling in mice. Am J Physiol Lung Cell Mol Physiol. Published ahead of print on October 15, 2004 as 10.1152/ajplung.00304.2004.
Broccard AF, Hotchkiss JR, Kuwayama N, Olson DA, Jamal S, Wangensteen DO, Marini JJ. Consequences of vascular flow on lung injury induced by mechanical ventilation. Am J Respir Crit Care Med 1998;157:1935eC1942.
Broccard AF, Vannay C, Feihl F, Schaller MD. Impact of low pulmonary vascular pressure on ventilator-induced lung injury. Crit Care Med 2002;30:2183eC2190.
Zeltner TB, Caduff JH, Gehr P, Pfenninger J, Burri PH. The postnatal development and growth of the human lung: I. Morphometry. Respir Physiol 1987;67:247eC267.
Yano Y, Braimbridge MV, Hearse DJ. Protection of the pediatric myocardium: differential susceptibility to ischemic injury of the neonatal rat heart. J Thorac Cardiovasc Surg 1987;94:887eC896.
Alessandrini F, D'Armini AM, Roberts CS, Reddick RL, Egan TM. When does the lung die II. Ultrastructural evidence of pulmonary viability after "death." J Heart Lung Transplant 1994;13:748eC757.
Strand M, Jobe AH. The multiple negative randomized controlled trials in perinatology—why Semin Perinatol 2003;27:343eC350.(Alik Kornecki, Shinya Tsu)
Departments of Anesthesia, Pediatrics, Physiology
the Interdepartmental Division of Critical Care Medicine, University of Toronto, Toronto
Intestinal Disease Research Program, Department of Medicine, McMaster University, Hamilton
Department of Pediatrics (Critical Care Unit, Children's Hospital of Western Ontario), University of Western Ontario, London, Ontario, Canada
ABSTRACT
Rationale: Ventilator-induced lung injury has been predominantly studied in adults. Objectives: To explore the effects of age and lung development on susceptibility to such injury. Methods: Ex vivo isolated nonperfused rat lungs (infant, juvenile, and adult) were mechanically ventilated where VT was based on milliliters per kilogram of body weight or as a percentage of the measured total lung capacity (TLC). In vivo anesthetized rats (infant, adult) were mechanically ventilated with pressure-limited VTs. Allocation to ventilation strategy was randomized. Measurements: Ex vivo injury was assessed by pressureeCvolume analysis, reduction in TLC, and histology, and in vivo injury by lung compliance, cytokine production, and wet- to dry-weight ratio. Main Results: Ex vivo ventilation (VT 30 ml · kgeC1) resulted in a significant reduction (36.0 ± 10.1%, p < 0.05) in TLC in adult but not in infant lungs. Ex vivo ventilation (VT 50% TLC) resulted in a significant reduction in TLC in both adult (27.8 ± 2.8%) and infant (10.6 ± 7.0%) lungs, but more so in the adult lungs (p < 0.05); these changes were paralleled by histology and pressureeCvolume characteristics. After high stretch in vivo ventilation, adult but not infant rats developed lung injury (total lung compliance, wet/dry ratio, tumor necrosis factor ). Surface video microscopy demonstrated greater heterogeneity of alveolar distension in ex vivo adult versus infant lungs. Conclusion: These data provide ex vivo and in vivo evidence that comparable ventilator settings are significantly more injurious in the adult than infant rat lung, probably reflecting differences in intrinsic susceptibility or inflation pattern.
Key Words: infant lung injury mechanical ventilation pediatric ventilator-induced lung injury
Lung injury, or worsened lung injury during mechanical ventilation, is clearly a matter of major concern. Clinicians caring for ventilated patients, as well as researchers investigating the responsible mechanisms, attempt to understand the optimal approach to ventilation to lessen lung injury. A current approach whereby excessive tidal volumes (VTs) are avoided (1, 2) has evolved over time on the basis of laboratory studies performed largely in adult animals (3). Furthermore, the demonstration that the manner of application of mechanical ventilation directly affects survival attests to the importance of this area of research and has been validated in a landmark randomized clinical trial in adults (4).
Compared with adults, pediatric patients demonstrate a spectrum of lung development spanning neonatal, infant, youth, and adult stages. Beyond neonatal age, there are important differences between infant and adult lungs (e.g., alveolar structure, matrix composition, angiogenesis) (5). In fact, maturation in the human lung continues well after the newborn period until between the ages of 2 and 8 years (5). Although lung injury in preterm neonates has been extensively investigated and characterized (6), there has been limited laboratory investigation (7eC9) and no prospective clinical investigation of ventilator-associated lung injury affecting infancy or youth. Instead, clinical research concerning mechanical ventilation in infants and young children (nonneonatal, nonadult) has focused not on VT or airway pressure but on adjuvant therapies, including inhaled nitric oxide (10), surfactant (11), and the use of high-frequency oscillation (12). As a result, evidence-based age-specific guidelines for the use of conventional mechanical ventilation in pediatric patients have not been possible; furthermore, any recommendations that exist have been extrapolated from adult data (13, 14).
An empiric sense that infant lungs might be more susceptible to ventilator-induced lung injury (VILI) compared with adult lungs is supported by a single laboratory investigation that compared juvenile with adult rabbits (7). That report, which demonstrated greater injury in the lungs of younger animals, concluded that the findings may have reflected the larger VTs delivered because of greater respiratory system compliance (7). However, there is an important reason to believe that infant lungs may be less susceptible to VILI. VT is usually expressed in terms of milliliters per kilogram of body weight, but the ratio of lung volume compared with body weight varies with development. In rats, it increases (15), and in humans, it may decrease (16). Rather than attempt to define locally applied mechanical stress, the current study evaluated the pulmonary responses to comparable levels of VT on the basis of body weight (ml · kgeC1) and lung volume (ml/ml, % total lung capacity [TLC]), as well as on inflation pressure.
Our main objective in this study was to explore whether lung maturation beyond the neonatal age range has an effect on susceptibility to VILI in the rat lung. The rat was chosen because of the well-characterized developmental morphology (15, 17, 18), which, after 4 days of age, is comparable to human lung development (19, 20). To remove any influence of age-related chest wall compliance or lungeCheart interaction, an ex vivo, isolated, nonperfused rat lung model was studied (21, 22). We then reproduced our experiments in the in vivo anesthetized rat. Some of the results of these studies have been previously reported in the form of an abstract (23).
METHODS
Male Sprague-Dawley rats of three maturity levels: adult (250eC360 g, age 75 days), juvenile (72eC100 g, age 35 days), and infant (26eC45 g, age 17 days) were used. The study was conducted according to the guidelines of the Canadian Council for Animal Care and was approved by the Animal Care Committee of the Hospital for Sick Children. Complete details of the experimental protocol are provided on the online supplement. The animals were anesthetized and ventilated (24) with peak inspiratory pressure of 9 to 10 cm H2O, positive end-expiratory pressure of 1 cm H2O, inspired oxygen of 0.21, and respiratory rates of 45 (adult), 60 (young), and 80 (infant) breaths/minute. For ex vivo experiments, lungs and heart were removed en bloc and suspended in a warm, humidified chamber (21, 25, 26).
Experimental Outline
After determination of baseline TLC and pressureeCvolume (P-V) curves, lungs in Series 1 and 2 were randomly allocated to receive either no ventilation (control) or mechanical ventilation for 60 minutes. In Series 1, ex vivo lungs were ventilated with a VT of 30 ml · kgeC1, a positive end-expiratory pressure of 0, a respiratory rate of 40 minuteseC1, an FiO2 of 0.21, and an FiCO2 of 0.05 for 60 minutes. In Series 2, ex vivo lungs were ventilated with a VT set to 50% of baseline TLC. After the experiments, lung weight was recorded. In Series 3, anesthesia was maintained and the intact in vivo infant and adult animals ventilated with a peak inspiratory pressure of 20 or 30 cm H2O for 90 minutes, and compared with nonventilated control animals. In the in vivo experiments, respiratory rate was the same in all groups (34 minuteseC1), and added dead space was used to obtain comparable PaCO2 levels (35eC45 mm Hg) in each group. In Series 4, lungs were ventilated as in Series 2, and with a respiratory rate of 30 minuteseC1. Alveolar inflation and deflation were measured using inverted microscopy and analyzed offline. Static P-V curves were constructed, before and after ventilation (21, 27). Lungs were excluded from the experiments in the event of air leaks or technical problems with preparation. After in vivo experiments (Series 3), the decrement in quasi-static respiratory compliance, bronchoalveolar lavage concentration of tumor necrosis factor (TNF-), and the lung wet/dry weight were determined.
Morphologic Analysis
The left lung was isolated and inflation-fixed in formalin (inflation pressure 20 cm H2O for 48 hours) (22). An assessment was performed independently by two observers, who were blinded as to group allocation and the mean values reported (22).
Statistical Analysis
The data are expressed as mean ± SD. Statistical comparisons were performed with either analysis of variance followed by Student-Neumann-Keuls tests (Series 1), or with Student's t test (Series 2, 3). TNF- (Series 3) was analyzed using the nonparametric Kruskal-Wallis test, and post hoc analysis was with Dunn's method. Histology scores between observers were compared using interclass correlation coefficient (0.88, 95% confidence intervals, 0.817eC0.924). We considered differences significant where p was less than 0.05.
RESULTS
Experimental Series 1: Ex Vivo VT Determined by Body Weight
Baseline characteristics.
A total of 54 animals were selected; 50 were randomized, and 48 completed the protocol. Baseline variables are reported in Table 1. The TLC corrected for body weight was ranked as follows: infant > juvenile > adult (Table 1; p < 0.05). This pattern was conserved for baseline chord compliance, which was also corrected for body weight (Table 1; p < 0.05). The baseline peak inspiratory pressure, recorded 10 minutes after commencement of ventilation, was significantly higher in the adult group than in the other groups, with a rank order as follows: adult > juvenile infant lungs (Table 1; p < 0.05).
Impact of ventilation on lung mechanics.
After 60 minutes of ventilation with high VT, the TLC decreased significantly from the baseline values in the adult and juvenile groups, but not in the infant group (Figure 1). The TLC (expressed in ml · kgeC1) decreased from 39.2 ± 3.0 to 25.0 ± 3.6 ml · kgeC1, a reduction of 36.0 ± 10.1%, in the adult group (Figure 1; p < 0.05). In the juvenile group, the TLC decreased from 56.4 ± 4.7 to 47.5 ± 6.1 ml · kgeC1, a reduction of 15.4 ± 10.1% (Figure 1; p < 0.05). In the infant group, the TLC was not significantly reduced after ventilation (Figure 1; 79.8 ± 9.2 vs. 75.9 ± 7.4 ml · kgeC1, p = 0.32). The rank order of high VTeCinduced decrease in TLC was as follows: adult > juvenile infant (Figure 1; p < 0.05).
In addition to the TLC data, the complete P-V characteristics are presented for each group. Because the P-V characteristics change with age (28), the baseline and final P-V curves for each of the groups were expressed in the following three ways: pressure (cm H2O) versus uncorrected volume (ml; Figure 2A), pressure (cm H2O) versus volume per kilogram of animal body weight (ml · kgeC1; Figure 2B) (7), and as pressure (cm H2O) versus volume expressed as a percentage of the preventilation TLC (Figure 2C) (29). Comparison of these P-V curves demonstrates the following features, which are independent of how the inflation and deflation volumes are expressed. First, the P-V curves displayed postventilation downward shift, reflecting a reduction in chord compliance as follows: adult > juvenile > infant (Figure 2). Second, there was no inflection point demonstrable in any of the preventilation curves, and after ventilation, a lower inflection point developed in the adult lungs, but not in the infant or juvenile lungs (Figure 2).
Impact of ventilation on lung histology.
There was no difference in alveolar injury, expressed as hyaline membrane score, among the nonventilated (control) adult, juvenile, or infant lungs (Figure 3). The hyaline membrane score was greater in ventilated than in nonventilated lungs in the adult group only, and not in the juvenile or infant groups. In ventilated lungs, the rank order of hyaline membrane score was as follows: adult > juvenile infant groups (Figure 3; p < 0.05). There were no differences detected in hyaline membrane score between proximal and distal airspaces in any of the groups (data not presented). Airway injury scores were not different among the nonventilated (control) lungs from adult juvenile or infant animals (Figure 4). The airway injury score was significantly greater in ventilated lungs compared with nonventilated lungs in adult, but not in juvenile or infant groups (Figure 4; p < 0.05). In ventilated lungs, the rank order of airway injury score was as follows: adult > juvenile infant groups (Figure 4; p < 0.05). There were no differences detected in airway injury score between proximal and distal airways (data not presented).
Series 2: Ex Vivo VT Determined by TLC
In Series 2, VT was dictated not by body weight (i.e., ml · kgeC1) but by the baseline lung volume (i.e., VT administered was 50% of baseline TLC). Pilot studies indicated that this approach was associated with lower peak airway pressures in the adult versus infant lungs. The greatest differences in Series 1 were between the adult and the infant lungs, and so these groups were the focus of Series 2.
Baseline characteristics.
In this series, seven adult and seven infant animals were selected; six from each group were studied, and all completed the experiments. The baseline variables are reported in Table 2.
Impact of ventilation on lung mechanics.
Despite the higher VT (ml · kgeC1) and peak inspiratory pressure in the infant lungs, 60 minutes of mechanical ventilation with a VT at 50% of baseline TLC was associated with a significantly smaller reduction in TLC in the infant versus the adult lungs (Figure 5; p < 0.05). In the adult lungs, TLC corrected for body weight decreased from 39.8 ± 2.88 to 28.5 ± 2.82 ml · kgeC1 (Figure 5; p < 0.05), whereas in the infant group, the TLC decreased from 68.3 ± 10.5 to 61.1 ± 10.9 ml · kgeC1 (Figure 5; p < 0.05). The percentage of reduction in TLC (% baseline value) was greater in the adult versus the infant lungs (28.7 ± 2.84 vs. 10.6 ± 7.0%; Figure 5; p < 0.05). This pattern was maintained with expression of the P-V curves in terms of the following: pressure (cm H2O) versus uncorrected volume (ml), pressure (cm H2O) versus volume corrected for body weight (ml · kgeC1), and as pressure (cm H2O) versus volume expressed as a percentage of the baseline volume at the same static pressure (Figure 6).
Impact of ventilation on lung histology.
Hyaline membrane scores and airway injury scores were both significantly greater after ventilation in the adult versus the infant lungs (Figure 7; see Figure E1 in the online supplement). No significant differences in hyaline membrane scores or airway injury score were observed between proximal and distal areas (data not presented).
Series 3: In Vivo Ventilation
Baseline characteristics.
Thirty-two animals were selected, and all completed the experiment. The baseline variables are reported in Table 3. When ventilated with 20 cm H2O, the resultant VTs in the infants versus adults were 37.5 ± 1.2 versus 18.4 ± 1.3 ml · kgeC1, respectively (Table 3; p < 0.05). Similarly, when ventilated with 30 cm H2O, the resultant VTs in the infants versus adults were 42.4 ± 4.4 compared with 29.7 ± 3.8 ml · kgeC1 (Table 3; p < 0.05).
Impact on lung mechanics and edema.
After 90 minutes of mechanical ventilation, no change in total lung compliance was detected in the adult animals ventilated with a pressure of 20 cm H2O, but a significant reduction in lung compliance occurred in those ventilated with a pressure of 30 cm H2O (Figure 8). Despite being ventilated with comparatively higher VTs (in terms of ml · kgeC1), no reductions in lung compliance occurred in any of the infant animals, whether ventilated with 20 or 30 cm H2O (Figure 8). The magnitude of the effect of ventilation on compliance in the adult and infant animals paralleled the changes in lung wet/dry ratio (Figure 9).
Impact on pulmonary TNF-.
The volume of bronchoalveolar lavage fluid retrieved was similar within each age group. An illustrative proinflammatory cytokine, TNF- (30), was not detected in nonventilated control animals and was significantly greater in ventilated adults ventilated with 30 cm H2O versus all other groups, with no other among-group differences (Figure 10; p < 0.05).
Series 4: Ex Vivo Ventilation—Lung Imaging
Three animals from each group (infant 43 ± 1.8 g, adult 331 ± 33 g) completed the experiment. The mean TLC (infant 38.2 ± 3.2 ml · kgeC1, adult 69 ± 7.2 ml · kgeC1) was similar to the respective values in Series 1 and 2. Mean alveolar diameter at the end-expiration was 39.7 ± 8.4 e in adults compared with a mean of 35.3 ± 4.5 e in infants (p = not significant). Mean alveolar diameter at end-inspiration was significantly greater in the adult versus infant lungs (77.3 ± 6.7 vs. 57.3 ± 2.0 e, p < 0.05). The mean variance of the alveolar diameter at the end of inspiration was significantly greater in adult versus infant lungs (824 ± 240 vs. 318 ± 40, p < 0.05; Figure 11).
DISCUSSION
The principal finding of this study is that, whether VT is based on body weight (ml · kgeC1) or baseline TLC (%), comparable ventilator settings resulted in greater injury in adult versus infant (nonneonatal, nonadult) lungs. The lessened lung injury in infants was confirmed in vivo, despite exposure to larger VT (relative to body weight) and higher peak inspiratory pressure. In addition, adult lungs appeared to have a more heterogeneous distribution of alveolar ventilation. Although the current study did not directly address the issues of local lung stress or strain (31), the overall data may suggest that intrinsic properties of the lung are responsible for the lesser susceptibility in infants.
Development and Susceptibility to Lung Injury
Laboratory or clinical studies of VILI have been performed mostly in surfactant-deficient preterm newborns (32, 33) or in adults (3, 4, 22, 34). Two articles have recently reported important findings that may relate to mechanisms of VILI in nonpremature newborn models (35, 36). Newborn rats exposed to high VT demonstrate less cytokine mRNA production compared with adult animals (36). In addition, Rau and coworkers (35) have demonstrated that, in a porcine model, the newborn has different surfactant composition—and superior dynamic function—than the adolescent.
Such comparative studies are important because they recognize the potential impact of development on susceptibility to VILI between newborn versus adult lungs. However, beyond the newborn or neonatal period, only three laboratory studies have addressed VILI in young animals (7eC9). The only study (7) that directly compared juveniles with adults demonstrated that, with comparably elevated peak inspiratory pressures, younger animals developed more lung injury. The authors speculated that, because of greater respiratory system compliance, the younger animals might have been exposed to disproportionately larger VT (7).
The current study focuses on the infant versus adult susceptibility, where infancy in this model is characterized by almost complete alveolarization (i.e., having attained the adult number of alveoli) (15, 18), and with the principal differences between infant (nonneonate) and adult lungs being the degree of matrix development (28, 37).
Choice of VT
Where VT is considered in the clinical context, it is usually specified as milliliters per kilogram of body weight. Although the relationship between VT, body weight, and ventilator-associated lung injury is complex in adults (38), there is an additional layer of complexity in children. This is because even where weights can be calculated with high degrees of accuracy, the ratio of lung volume (TLC) to body weight changes with age and development. The ratio of TLC to body weight in the present study decreased over maturation from 74.0 ± 8.8 ml · kgeC1 in infant, to 53.4 ± 5.6 ml · kgeC1 in juvenile, and to 39.7 ± 4.0 ml · kgeC1 in adult groups (data averaged from Tables 1 and 2). These findings are comparable to previously reported data for the developing rat (15, 28, 39, 40). Although the optimal measurement of lung volume in young children is not established, the available data suggest that TLC/body weight ratio increases with age from 52 ± 13 ml · kgeC1 in infants to 87 ± 11 ml · kgeC1 in older children (16, 29).
Potential Mechanisms
There are several possible explanations for the current findings. The absolute VT was different among the groups. However, in Series 1, the VT was identical in terms of milliliters per kilogram of body weight and in Series 2 was dictated by TLC (50% TLC). In addition, airway pressure, although slighter higher in the adult group in Series 1, was significantly lower ( 2 cm H2O) in the adult group in Series 2. Thus, neither differences in VT nor in airway pressure explain the among-group differences in injury; in fact, the study was biased against finding worse injury in the adult lungs in Series 2, because both the VT (in ml · kgeC1) and airway pressures were lower in the adult lungs.
The inclusion of in vivo experiments provides additional important information, because the presence of the intact chest wall, pulmonary perfusion with intact blood, as well as the potential for cardiopulmonary interactions corroborate the ex vivo findings. There are significant differences in baseline TLC, which, when corrected for body weight, are far larger in infant versus adult lungs (Tables 1 and 2). This result could certainly explain the findings in Series 1 where VT based on body weight would occupy a smaller fraction of TLC in infants versus adults and therefore cause less stretch-induced injury, but this could not explain the differences observed in Series 2. Lung compliance represents the interaction of pressure and volume and determines how a given VT alters pressure and vice versa. However, comparing baseline compliance among different age groups is complex, and although correction for either body weight or lung weight has been reported (41), there is no universally agreed-on approach. Both approaches have been used in the current experiments, and when corrected for body weight, the compliance is greater in infants, but when corrected for lung weight, is greater in adults. Overall, the complexity of correcting or normalizing lung parameters for either body weight or lung weight is apparent from the current study, as well as from previous publications (15, 28, 39eC41). It seems logical therefore to expect a different impact of "body weighteCdirected" versus "lung volumeeCdirected" prescription of VT.
The pattern of alveolar distension might provide additional explanation. Real-time direct visualization of the lung has been recently described (42), and has provided important insights into alveolar micromechanics. Notwithstanding the sampling limitations of such surface visualization of subpleural alveoli (43), the current study describes a greater degree of heterogeneous subpleural alveolar inflation in adult versus infant lungs. This finding raises, but does not prove, the possibility that either underdistension (22) and/or overdistension (44) may account for the differences in injury between the two groups. Furthermore, the downward and rightward shift of the adult P-V curves after ventilation, as well as the appearance of a lower inflection point, may support the previously described findings and suggest an atelectasis-associated mechanism of injury in the adult lung (Figures 2 and 6) (45).
It is possible that the infant lung, which at 17 days has almost full alveolarization but with smaller alveoli, has an increased quantity of surfactant relative to the alveolar size. If true, this could play a role in the increased uniformity of inflation and lesser susceptibility to VILI observed in the infant lungs. Unfortunately, surfactant analysis was not performed in the current study.
An integrated explanation of these phenomena may involve appreciation of the differences in the matrix of infant versus adult lungs. The levels of collagen and elastin, the major matrix components, undergo significant changes between 4 and 40 days (28). Collagen concentration increases linearly over the interval from infancy to adulthood, whereas elastin concentration increases 10-fold over the first 20 days and rises less rapidly thereafter (46). In addition, the cross-linking of collagen changes with age. Furthermore, data from lung slice experiments indicate that mechanical stretch changes lung matrix properties (47). Thus, it is possible that the age-related change in ratio of elastin to collagen could contribute to altered susceptibility to stretch-induced lung injury. Finally, the recent studies cited previously (35, 36) may provide additional explanation. If the relatively blunted cytokine response (36) or enhanced surfactant function (35) that were reported in the newborn are also present in older infants, then the lowered susceptibility to ventilator-associated lung injury of infants may be, in part, attributable to these phenomena.
Pulmonary Cytokine Release
The role of cytokines in general and TNF- in particular in the genesis of VILI is controversial (48, 49). The observation that the concentration of TNF- increases significantly after high VT ventilation in vivo in the adult but not the infant (Figure 10) mirrors a previous report (36) where mechanical ventilation promoted less production of cytokine in newborn compared with adult rats. That finding may be explained by a difference in immunologic maturation between adult versus infant rats, or by the known differences in the matrix composition that may affect the way that global lung stretch is transmitted to individual cells.
Choice of Experimental Models and Study Limitations
The use of an ex vivo preparation alone would limit the applicability of the data. The current data, however, include in vivo confirmation, and the ex vivo data control for the impact of pulmonary vascular pressure or flow on injury (50, 51). The TLC in this study was measured in nondegassed lung. In fact, we measured the volume above the FRC volume. Furthermore, the value of FRC may differ across progressive stages of lung development, and an increased ratio of FRC to TLC has been described in newborn versus adult rats (41). We know of no data regarding true FRC values in the isolated nonperfused rat lung at infancy, but (at any age) they are almost certainly far smaller than in vivo values. Because each lung served as its own control, baseline differences in FRC among the groups should not alter the conclusions.
All animal models of human development are necessarily limited. Although there are important differences in development between the human and rat lung (52), the rat lung provides a good model for human lung development after the age of 4 days (19). At 1 month, the human lung is morphologically comparable to a rat lung aged 1 week, and at 2 to 5 years, the human lung is comparable to a rat lung aged 17 to 21 days (20). It is possible that development confers differences in viability on ex vivo lung preparations. Furthermore, data indicate that infants may be less susceptible than adults to organ ischemia or reperfusion (53). However, the current ex vivo experiments were performed within the known limits of lung viability (54), and such issues would not affect the in vivo data.
We have tested three different sets of comparisons (i.e., VT according to body weight, TLC, or by pressure); however, the current study did not directly address the concepts of stress or strain at the local tissue level. The central nature of these concepts has been commented on previously (31). Lung weight was measured in Series 2, and the higher ratio of VT (ml) to lung weight (g) in the adult versus the infant lungs raises the possibility that stress per unit of lung tissue mass might contribute to the increased injury in the adult lungs (Table 2).
Because neither airway resistance nor plateau pressure were recorded, it is possible that the lesser injury observed in the infant may be attributed to higher airway resistance or to auto-positive end-exipirator pressure. However, as previously suggested by Gomes and coworkers (40), airway resistance in rats, when corrected for weight, is not significantly different between the infant and adult. In addition, the length of the endotracheal tube was designed to abolish any differences in endotracheal tube resistance (see online supplement for details). Furthermore, from the clinical perspective, lung mechanical stretch is extrapolated from the magnitude of VT and/or plateau pressure. Thus, inclusion of plateau pressure, especially in the ex vivo preparations, should be considered in future work.
Potential Significance
This study, the first study to make age-specific comparisons of VILI between infants and adults, suggests four potential lessons. First, infant lungs may be less susceptible to VILI when exposed to comparable ventilation settings of VT or pressure. Second, development or propagation of stretch may be different in infant than adult lungs. Furthermore, if the differences observed in inflation pattern (Figure 11) are confirmed in other studies, then the utility of positive end-expiratory pressure or recruitment maneuvers may be different in infants versus adults. Third, we suggest caution in extrapolating recommendations about ventilator settings from adult studies to children, especially where the VT is prescribed on the basis of milliliters per kilogram of body weight. Finally, because understanding preclinical phenomena is an important factor in the design of successful clinical trials (55), studies such as the current report might eventually influence clinical evaluation of pediatric ventilatory strategies that are distinct from adult studies.
Acknowledgments
The authors thank Dr. A.C. Bryan for his insightful comments and Derek Stephens (Program in Population Health Sciences, Hospital for Sick Children) for statistical expertise.
This article has an online supplement, which is accessible from this issue's table of contents at www.atsjournals.org
REFERENCES
Marini JJ, Gattinoni L. Ventilatory management of acute respiratory distress syndrome: a consensus of two. Crit Care Med 2004;32:250eC255.
Tobin MJ. Advances in mechanical ventilation. N Engl J Med 2001;344:1986eC1996.
Dreyfuss D, Saumon G. Ventilator-induced lung injury: lessons from experimental studies. Am J Respir Crit Care Med 1998;157:294eC323.
Acute Respiratory Distress Syndrome Network. Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung injury and the acute respiratory distress syndrome. N Engl J Med 2000;342:1301eC1308.
Chernick V, Boat TF, Kendig EL, editors. Kending's disorders of the respiratory tract in children, 6th ed. Philadelphia: W.B. Saunders; 1998.
Speer CP. Inflammatory mechanisms in neonatal chronic lung disease. Eur J Pediatr 1999;158:S18eCS22.
Adkins WK, Hernandez LA, Coker PJ, Buchanan B, Parker JC. Age effects susceptibility to pulmonary barotrauma in rabbits. Crit Care Med 1991;19:390eC393.
Peevy KJ, Hernandez LA, Moise AA, Parker JC. Barotrauma and microvascular injury in lungs of nonadult rabbits: effect of ventilation pattern. Crit Care Med 1990;18:634eC637.
Coker PJ, Hernandez LA, Peevy KJ, Adkins K, Parker JC. Increased sensitivity to mechanical ventilation after surfactant inactivation in young rabbit lungs. Crit Care Med 1992;20:635eC640.
Okamoto K, Hamaguchi M, Kukita I, Kikuta K, Sato T. Efficacy of inhaled nitric oxide in children with ARDS. Chest 1998;114:827eC833.
Dekowski SA, Holtzman RB. Surfactant replacement therapy: an update on applications. Pediatr Clin North Am 1998;45:549eC572.
Hoehn T, Krause M, Krueger M, Hentschel R. Treatment of respiratory failure with inhaled nitric oxide and high-frequency ventilation in an infant with respiratory syncytial virus pneumonia and bronchopulmonary dysplasia. Respiration (Herrlisheim) 1998;65:477eC480.
Matthews BD, Noviski N. Management of oxygenation in pediatric acute hypoxemic respiratory failure. Pediatr Pulmonol 2001;32:459eC470.
Hirschl RB. Support of respiratory failure in the pediatric surgical patient. Curr Opin Pediatr 2002;14:459eC469.
Burri PH, Dbaly J, Weibel ER. The postnatal growth of the rat lung: I. Morphometry. Anat Rec 1974;178:711eC730.
Thorsteinsson A, Larsson A, Jonmarker C, Werner O. Pressure-volume relations of the respiratory system in healthy children. Am J Respir Crit Care Med 1994;150:421eC430.
Burri PH. Postnatal growth and maturation of the lung. Chest 1975;67:2SeC3S.
Burri PH. The postnatal growth of the rat lung: 3. Morphology. Anat Rec 1974;180:77eC98.
Cherukupalli K, Larson JE, Puterman M, Sekhon HS, Thurlbeck WM. Comparative biochemistry of gestational and postnatal lung growth and development in the rat and human. Pediatr Pulmonol 1997;24:12eC21.
Zeltner TB, Burri PH. The postnatal development and growth of the human lung: II. Morphology. Respir Physiol 1987;67:269eC282.
Tremblay L, Valenza F, Ribeiro SP, Li J, Slutsky AS. Injurious ventilatory strategies increase cytokines and c-fos m-RNA expression in an isolated rat lung model. J Clin Invest 1997;99:944eC952.
Muscedere JG, Mullen JB, Gan K, Slutsky AS. Tidal ventilation at low airway pressures can augment lung injury. Am J Respir Crit Care Med 1994;149:1327eC1334.
Kornecki A, Tsuchida S, Engelberts D, Frndova H, Belik J, Kavanagh BP. Ventilator-associated lung injury: are infants as susceptible as adults Am J Respir Crit Care Med 2004;169:A787.
Duggan M, McCaul CL, McNamara PJ, Engelberts D, Ackerley C, Kavanagh BP. Atelectasis causes vascular leak and lethal right ventricular failure in uninjured rat lungs. Am J Respir Crit Care Med 2003;167:1633eC1640.
Veldhuizen RA, Tremblay LN, Govindarajan A, van Rozendaal BA, Haagsman HP, Slutsky AS. Pulmonary surfactant is altered during mechanical ventilation of isolated rat lung. Crit Care Med 2000;28:2545eC2551.
Ribeiro SP, Rhee K, Tremblay L, Veldhuizen R, Lewis JF, Slutsky AS. Heat stress attenuates ventilator-induced lung dysfunction in an ex vivo rat lung model. Am J Respir Crit Care Med 2001;163:1451eC1456.
Venegas JG, Harris RS, Simon BA. A comprehensive equation for the pulmonary pressure-volume curve. J Appl Physiol 1998;84:389eC395.
Nardell EA, Brody JS. Determinants of mechanical properties of rat lung during postnatal development. J Appl Physiol 1982;53:140eC148.
Fagan DG. Post-mortem studies of the semistatic volume-pressure characteristics of infants' lungs. Thorax 1976;31:534eC543.
Chiumello D, Pristine G, Slutsky AS. Mechanical ventilation affects local and systemic cytokines in an animal model of acute respiratory distress syndrome. Am J Respir Crit Care Med 1999;160:109eC116.
Marini JJ. Relative importance of stretch and shear in ventilator-induced lung injury. Crit Care Med 2004;32:302eC304.
Clark RH, Gerstmann DR, Jobe AH, Moffitt ST, Slutsky AS, Yoder BA. Lung injury in neonates: causes, strategies for prevention, and long-term consequences. J Pediatr 2001;139:478eC486.
Bhuta T, Henderson-Smart DJ. Elective high-frequency oscillatory ventilation versus conventional ventilation in preterm infants with pulmonary dysfunction: systematic review and meta-analyses. Pediatrics 1997;100:E6.
Dreyfuss D, Soler P, Basset G, Saumon G. High inflation pressure pulmonary edema: respective effects of high airway pressure, high tidal volume, and positive end-expiratory pressure. Am Rev Respir Dis 1988;137:1159eC1164.
Rau GA, Vieten G, Haitsma JJ, Freihorst J, Poets CF, Ure BM, Bernhard W. Surfactant in newborn compared to adolescent pigs: adaptation to neonatal respiration. Am J Respir Cell Mol Biol 2003;30:694eC701.
Copland IB, Martinez F, Kavanagh BP, Engelberts D, McKerlie C, Belik J, Post M. High tidal volume ventilation causes different inflammatory responses in newborn versus adult lung. Am J Respir Crit Care Med 2004;169:739eC748.
Tanaka R, Ludwig MS. Changes in viscoelastic properties of rat lung parenchymal strips with maturation. J Appl Physiol 1999;87:2081eC2089.
Thompson TB, Hayden D, Matthay MA, Brower R, Parsons PE. Clinicians' approaches to mechanical ventilation in acute lung injury and ARDS. Chest 2001;120:1622eC1627.
Bartlett D Jr, Areson JG. Quantitative lung morphology in newborn mammals. Respir Physiol 1977;29:193eC200.
Gomes RF, Shardonofsky F, Eidelman DH, Bates JH. Respiratory mechanics and lung development in the rat from early age to adulthood. J Appl Physiol 2001;90:1631eC1638.
Fisher JT, Mortola JP. Statics of the respiratory system in newborn mammals. Respir Physiol 1980;41:155eC172.
Steinberg J, Schiller HJ, Halter JM, Gatto LA, Dasilva M, Amato M, McCann UG, Nieman GF. Tidal volume increases do not affect alveolar mechanics in normal lung but cause alveolar overdistension and exacerbate alveolar instability after surfactant deactivation. Crit Care Med 2002;30:2675eC2683.
Kavanagh BP. Lung recruitment in real time: learning was never so easy. Am J Respir Crit Care Med 2003;167:1585eC1586.
Dreyfuss D, Saumon G. Barotrauma is volutrauma, but which volume is the one responsible Intensive Care Med 1992;18:139eC141.
Martin-Lefevre L, Ricard JD, Roupie E, Dreyfuss D, Saumon G. Significance of the changes in the respiratory system pressure-volume curve during acute lung injury in rats. Am J Respir Crit Care Med 2001;164:627eC632.
Brody JS, Therlbeck WM. Development, growth, and aging of the lung. In: Fishman MP, editor. Handbook of physiology: the respiratory system—mechanics of breathing. Vol. 3. Bethesda, MD: American Physiological Society; 1986. pp. 355eC386.
Al-Jamal R, Ludwig MS. Changes in proteoglycans and lung tissue mechanics during excessive mechanical ventilation in rats. Am J Physiol Lung Cell Mol Physiol 2001;281:L1078eCL1087.
Dreyfuss D, Ricard JD, Saumon G. On the physiologic and clinical relevance of lung-borne cytokines during ventilator-induced lung injury. Am J Respir Crit Care Med 2003;167:1467eC1471.
Wilson MR, Choudhury S, Takata M. Pulmonary inflammation induced by high stretch ventilation is mediated by tumor necrosis factor signalling in mice. Am J Physiol Lung Cell Mol Physiol. Published ahead of print on October 15, 2004 as 10.1152/ajplung.00304.2004.
Broccard AF, Hotchkiss JR, Kuwayama N, Olson DA, Jamal S, Wangensteen DO, Marini JJ. Consequences of vascular flow on lung injury induced by mechanical ventilation. Am J Respir Crit Care Med 1998;157:1935eC1942.
Broccard AF, Vannay C, Feihl F, Schaller MD. Impact of low pulmonary vascular pressure on ventilator-induced lung injury. Crit Care Med 2002;30:2183eC2190.
Zeltner TB, Caduff JH, Gehr P, Pfenninger J, Burri PH. The postnatal development and growth of the human lung: I. Morphometry. Respir Physiol 1987;67:247eC267.
Yano Y, Braimbridge MV, Hearse DJ. Protection of the pediatric myocardium: differential susceptibility to ischemic injury of the neonatal rat heart. J Thorac Cardiovasc Surg 1987;94:887eC896.
Alessandrini F, D'Armini AM, Roberts CS, Reddick RL, Egan TM. When does the lung die II. Ultrastructural evidence of pulmonary viability after "death." J Heart Lung Transplant 1994;13:748eC757.
Strand M, Jobe AH. The multiple negative randomized controlled trials in perinatology—why Semin Perinatol 2003;27:343eC350.(Alik Kornecki, Shinya Tsu)