Abstract
Background
Elevated surfactant protein D (SP-D) is a relatively specific indicator of lung injury and is associated with both acute and chronic lung disease in adults and respiratory distress syndrome in premature infants. The relationship between plasma SP-D and lung injury in children with acute respiratory failure is unclear.
Research Question
Is plasma SP-D associated with lung injury or outcome in children with acute respiratory failure?
Study Design and Methods
This was a prospective cohort study in children 2 weeks to 17 years of age with acute respiratory failure who participated in the BALI multi-center study. Analyses were done using SP-D levels in plasma from the first sample taken on either the day of intubation or one of the following 2 days. SP-D level was measured by enzyme-linked immunosorbent assay.
Results
Plasma samples from 350 patients were used in the analysis; 233 had pediatric ARDS (PARDS). SP-D levels varied across primary diagnoses (P < .001). Elevated SP-D levels were associated with severe PARDS after adjusting for age, pediatric risk of mortality III (PRISM-III), and primary diagnosis (OR = 1.02; CI = 1.01-1.04; P = .011). Multivariable analyses also indicated that elevated SP-D levels were associated with death (OR = 1.02; CI = 1.01-1.04; P = .004), duration of mechanical ventilation (P = .012), PICU length of stay (P = .019), and highest oxygenation index (P = .040). SP-D levels also correlated with age (rs = 0.16, P = .002).
Interpretation
Elevated plasma SP-D levels are associated with severe PARDS and poor outcomes in children with acute respiratory failure. Future studies will determine whether SP-D can be used to predict the degree of lung injury or response to treatment and whether SP-D is useful in identifying PARDS endotypes.
Key Words: ARDS, biomarker, critical illness, length of mechanical ventilation, mortality, outcome, pediatric ARDS, pediatrics, surfactant
Abbreviations: IQR, interquartile range; OI, oxygenation index; OSI, oxygen saturation index; PARDS, pediatric ARDS; PRISM-III, pediatric risk of mortality III; SP-D, surfactant protein D
FOR EDITORIAL COMMENT, SEE PAGE 850
Surfactant protein D (SP-D) is a protein first found in surfactant and later identified as a secretory product of type II alveolar epithelial cells.1 SP-D is constitutively expressed in type II alveolar cells, and levels are upregulated in response to lung injury or infection.2,3 SP-D is a member of the collectin family and is part of the host’s innate immune defense system. SP-D binds to oligosaccharides on the surface of a number of microorganisms, enhancing phagocytosis; it has also been reported to modulate other aspects of the innate immune response in the lung.2,4, 5, 6 Serum SP-D levels appear to be lower in children than in adults.7
Elevated SP-D in serum or plasma is associated with both acute and chronic lung injury in adults and is considered to be a relatively specific indicator of injury in the lung.6,8,9 Adult studies have shown that SP-D is associated with ARDS10, 11, 12 and with poor outcome in patients with ARDS.8,13 In pediatrics, elevated serum SP-D has been reported to be associated with respiratory distress syndrome in neonates.14 In addition, one small study15 reported that plasma SP-D is elevated in children with ARDS (n = 18); however, multivariable analysis indicated that SP-D was not associated with duration of mechanical ventilation. Because of the size of the study on children with ARDS, whether SP-D levels are associated with ARDS or poor outcomes in children with acute respiratory failure remains unclear. Nor is it clear whether the level of SP-D is indicative of the severity of lung injury in children with acute respiratory failure. We have now examined these questions in a large cohort of children with acute respiratory disease from the Genetic Variation and Biomarkers in Children with Acute Lung Injury (BALI) study.16 We hypothesized that elevated levels of plasma SP-D would be associated with pediatric ARDS (PARDS) and poor outcomes in children with acute respiratory failure.
Materials and Methods
This study, Genetic Variation and Biomarkers in Children with Acute Lung Injury (BALI), was an ancillary study to the multi-site clinical trial Randomized Evaluation of Sedation Titration for Respiratory Failure (RESTORE), which examined whether children mechanically ventilated for acute respiratory failure managed with a goal-directed sedation protocol had a decreased duration of mechanical ventilation compared with children receiving usual care.17 RESTORE enrolled children 2 weeks to 17 years old treated with invasive mechanical ventilation for acute airways or parenchymal lung disease between June 2009 and December 2013. Children for whom the length of mechanical ventilation was unlikely to be altered by the sedation management protocol being examined in RESTORE (ie, children who were ventilator dependent on PICU admission or those expected to be extubated within 24 h) were excluded. There were no additional inclusion or exclusion criteria for the BALI study. As reported previously, the demographic and clinical characteristics of children in BALI (n = 549) are very similar to those reported for patients enrolled in the RESTORE clinical trial (n = 2,449).16 BALI was designed in part to examine the association of specific plasma protein biomarkers with PARDS in prospectively enrolled children with acute respiratory failure. Details of the study methodology have been published previously.16 Twenty-two of the 31 PICUs participating in RESTORE also participated in BALI. The study was approved by the institutional review boards at all participating sites, as listed in e-Appendix 1. SP-D was measured in the first plasma sample taken after enrollment in the study, and measurements were included in the analysis if the first sample was taken on either the day of intubation (day 0), or on one of the following 2 days. Plasma SP-D was assayed in duplicate by an enzyme-linked immunosorbent assay (Yamasa Corporation) that is highly specific for SP-D (<0.3% cross-reactivity with human mannose binding lectin or surfactant protein A).18
The primary analyses examined the association of SP-D with the presence of PARDS. A priori power calculations using data from adults10 estimated that 180 patients with approximately 35% having PARDS would provide 90% power to see the difference in plasma SP-D levels observed in adults. PARDS was defined using oxygenation index (OI) or oxygen saturation index (OSI) as described by the Pediatric Acute Lung Injury Consensus Conference19 (PALICC), except that all patients defined as having PARDS also had bilateral infiltrates within 2 days before, or 1 day after, meeting OI or OSI criteria for PARDS. The PALICC definition for PARDS includes imaging findings consistent with acute pulmonary parenchymal disease, with stratification of patients into those with unilateral or bilateral infiltrates. BALI and RESTORE were designed when PARDS was defined using the American European Consensus Conference definition20; consequently, bilateral infiltrates were used to define PARDS because chest radiographic data from RESTORE only included the presence or absence of bilateral infiltrates. Secondary analyses examined association of SP-D with death (all-cause 90-day in-hospital mortality), duration of mechanical ventilation defined as done in RESTORE17 (patients assigned 28 days if they remained intubated or died before day 28, therefore making this outcome equivalent to ventilator-free days), PICU length of stay in survivors, and highest OI between days 0 and 2. If only the OSI was available, it was converted into an OI as described previously.21
Statistics
All analyses were completed using SAS 9.4. Basic descriptive analyses of demographic and key dependent variables (including frequency distributions or medians and interquartile ranges [IQR]) were conducted. Bivariate analyses were completed using Spearman correlations, Wilcoxon rank-sum, Kruskal-Wallis, χ2, or Fisher exact tests. If the Kruskal-Wallis tests were significant (P < .05), post hoc tests were run with the Dunn’s Multiple Comparison test.22 The results of the bivariate analyses were used to develop the final multivariable models. The potential covariates of age, sex, race and ethnicity, primary diagnosis, comorbid conditions, and severity of illness (PRISM-III score) were assessed for associations with outcomes or SP-D levels. Final models included variables (age, primary diagnosis, and PRISM-III score) that bivariate analysis indicated were significantly associated (P < .05) with outcomes or SP-D levels, with additional models that also included the day the sample was taken. Although some comorbid conditions and race/ethnicity were associated with SP-D, they were also associated with primary diagnosis, so we chose to include only primary diagnosis in the final models. Multivariable logistic regression models were used to examine the association of SP-D with the outcomes PARDS and mortality. Because of the low death rate, primary diagnosis could not be included in the mortality models; consequently, the primary diagnosis of asthma was included because levels of SP-D in children with a primary diagnosis of asthma appeared to be significantly lower than in most other diagnoses. Association of SP-D with duration of mechanical ventilation and PICU length of stay were examined using negative binomial regression adjusting for age, PRISM-III score, and primary diagnosis. Association of SP-D with highest OI was examined with multivariable linear regression adjusting for age, PRISM-III score, and primary diagnosis. Because the OI was not normally distributed, it was log transformed before use in the analysis.
Results
Of the 549 patients with acute respiratory failure enrolled in BALI, 350 had samples taken within 2 days of intubation. As shown in e-Table 1, the general demographic and clinical characteristics of the cohort with a plasma sample taken on days 0, 1, or 2 did not differ from those of the total cohort, with the exception that the median age of those with samples taken in the time frame of interest was slightly greater than that seen in the total cohort (4.6 y, IQR, 0.9-11.9; vs 3.8 y, IQR, 0.6-11.0; P < .01), and the proportion of patients from intervention sites is slightly higher (P = .04).
SP-D levels varied significantly across primary diagnoses (Fig 1; P < .0001), with SP-D levels in patients with a primary diagnosis of asthma significantly lower than in those with pneumonia, or those in the group designated as “Other” (P < .05). The diagnostic categories used in this analysis mirrored those used in the parent RESTORE trial, with the majority of patients in “Other” category (63%) having the disparate diagnoses of acute respiratory failure after bone marrow transplant (n = 8), thoracic trauma (n = 5), and acute chest syndrome (n = 4). SP-D levels in the “Other” group varied widely (e-Fig 1), with high levels of SP-D observed in patients with acute respiratory failure after bone marrow transplantation, thoracic trauma, acute chest syndrome, and pulmonary edema; when grouped together these patients had significantly higher SP-D levels than patients with pulmonary hemorrhage, croup or tracheitis, chronic lung disease, and multiple transfusions (median, IQR: 32.0, 17.5-54.4 ng/mL, n = 20 vs 4.8, 0-12.5 ng/mL, n = 7; P < .001). SP-D was weakly correlated with age in the whole group (n = 350, correlation coefficient [rs] = 0.16; P = .002), although the correlation strengthened when examined only in those with a primary diagnosis of pneumonia, as indicated by the higher correlation coefficient, (n = 124, rs = 0.30, P < .001).
Figure 1.
Surfactant protein-D levels across primary diagnoses. Plasma SP-D levels are significantly different across primary diagnoses (P < .0001, determined using Kruskal-Wallis test). Plasma SP-D is significantly higher in patients with pneumonia compared with those with a primary diagnosis of asthma and significantly higher in the “Other” group compared with those with a primary diagnosis of asthma or bronchiolitis. The diagnoses of those in the “Other” group and the corresponding median SP-D are shown in e-Figure 1. ∗P < .05, Dunn’s Multiple Comparisons test. The number within the bars is the number of patients in each group. SP-D = surfactant protein D.
The proportion of children with PARDS in the cohort was high, with 67% (n = 233/350) meeting the criteria for PARDS between days 0 and 2; 87% of those with PARDS met criteria on the day of intubation (day 0). A comparison of general demographic and clinical characteristics in those with and without PARDS in the group analyzed here indicated that the two groups were very similar, although both the frequency of having asthma as a primary diagnosis, or of having a medical history of asthma, was lower in children with PARDS (Table 1). Median SP-D in those with and without PARDS did not differ significantly (11.8 ng/mL, IQR = 6.6-20.8 ng/mL; 10.1 ng/mL, IQR = 6.0-16.2 ng/mL, respectively; P = .06, e-Fig 2). Multivariable logistic regression indicated that the association of SP-D with PARDS did not reach statistical significance (P = .06) when adjusting for age, PRISM-III, and primary diagnosis (Table 2).
Table 1.
Characteristics of Patients With and Without PARDS
| Characteristics | No PARDS n = 117 |
PARDS n = 233 |
P |
|---|---|---|---|
| No. (%) | |||
| Female | 55 (47) | 106 (45) | .79 |
| Non-Hispanic white | 62 (53) | 127 (55) | .73 |
| RESTORE intervention site | 75 (64) | 143 (62) | .65 |
| Primary diagnoses | |||
| Pneumonia | 35 (30) | 89 (38) | .13 |
| Bronchiolitis | 19 (16) | 41 (18) | .75 |
| Acute respiratory failure related to sepsis | 21 (18) | 51 (22) | .39 |
| Asthma or reactive airway disease | 20 (17) | 22 (9) | .04 |
| Aspiration pneumonia | 10 (9) | 15 (6) | .47 |
| Other | 12 (10) | 15 (6) | .21 |
| Medical history of | |||
| Prematurity | 10 (9) | 29 (12) | .27 |
| Asthma | 28 (24) | 36 (15) | .05 |
| Seizure disorder | 11 (9) | 21 (9) | .91 |
| Cancer | 9 (8) | 16 (7) | .78 |
| Neurologic/neuromuscular disorder | 12 (10) | 17 (7) | .34 |
| Chromosomal abnormality | 7 (6) | 23 (10) | .22 |
| Immunodeficiency | 3 (3) | 7 (3) | 1.0 |
| Deaths | 13 (11) | 14 (6) | .09 |
| PRISM III, median (IQR) | 7 (3-13) | 9 (5-14) | .08 |
| Age, median (IQR), y | 4.7 (0.8-11.1) | 4.6 (0.9-12.2) | .41 |
IQR = interquartile range; PARDS = pediatric ARDS; PRISM III = pediatric risk of mortality III.
Table 2.
Multivariable Analysis of Association of SP-D With PARDS or Severe PARDS in Children With Acute Respiratory Failure
| Variable | OR | 95% CI | P |
|---|---|---|---|
| PARDS | |||
| SP-D | 1.02 | 1.00, 1.03 | .059 |
| Age | 1.00 | 0.96, 1.05 | .913 |
| PRISM III | 1.04 | 1.01, 1.08 | .020 |
| Primary diagnosis | .095 | ||
| Pneumonia | Ref | ||
| Bronchiolitis | 1.10 | 0.53, 2.30 | .791 |
| Sepsis | 0.80 | 0.41, 1.56 | .511 |
| Asthma | 0.48 | 0.23, 1.00 | .051 |
| Aspiration Pneumonia | 0.54 | 0.21, 1.34 | .182 |
| Other | 0.35 | 0.14, 0.89 | .027 |
| Severe PARDS | |||
| SP-D | 1.02 | 1.01, 1.04 | .011 |
| Age | 1.04 | 1.00, 1.09 | .052 |
| PRISM III | 1.05 | 1.02, 1.08 | .004 |
| Primary diagnosis | .028 | ||
| Pneumonia | Ref | ||
| Bronchiolitis | 0.55 | 0.25, 1.22 | .140 |
| Sepsis | 0.56 | 0.30, 1.07 | .080 |
| Asthma | 0.26 | 0.10, 0.66 | .004 |
| Aspiration Pneumonia | 0.70 | 0.28, 1.76 | .446 |
| Other | 0.32 | 0.12, 0.88 | .027 |
N = 350 total; n = 233 with PARDS, n = 120 severe PARDS. ref = reference; SP-D = surfactant protein-D.
See Table 1 legend for expansion of other abbreviations.
To test the association of plasma SP-D levels and PARDS severity, patients were assigned to a severity level, as described in the PALICC definition,20 based on their worst OI or OSI (if OI was not available) across days 0 to 2. Comparison of median SP-D in those without PARDS and those with different PARDS severity levels indicated that the median level in children with severe PARDS was significantly higher than that seen in children with mild PARDS or without PARDS (Fig 2). Multivariable analysis adjusted for the same covariates described indicated a significant association of plasma SP-D with severe PARDS (Table 2).
Figure 2.
Surfactant protein-D levels in the absence of PARDS and across PARDS severities. Plasma SP-D levels are significantly different between patients without PARDS and different PARDS severity levels (P = .003 determined using using Kruskal-Wallis test). Plasma SP-D is significantly higher in patients with severe PARDS compared with those with no PARDS or mild PARDS (∗P < .05, Dunn’s Multiple Comparisons test). The number within the bars is the number of patients in each group. PARDS = pediatric ARDS. See Figure 1 legend for expansion of other abbreviation.
Because the first plasma sample could have been obtained on study day 0, 1, or 2, we also analyzed the impact of sample collection day on the association of SP-D with PARDS or severe PARDS. Day of sample collection was not associated with PARDS or severe PARDS, nor did its addition to the models impact the results of the multivariable analyses described previously (e-Table 2).
The association of SP-D with other clinical outcomes was also examined. SP-D was significantly higher in nonsurvivors (Fig 3). Higher SP-D levels were significantly associated with risk of death even after adjusting for age, PRISM-III score, and a diagnosis of asthma (Table 3). Adjusting for all primary diagnoses was not feasible because of the low mortality. SP-D was also associated with death (OR = 1.02; CI, 1.01-1.04; P = .005) when the sample day was added to the model, indicating that SP-D is associated with death irrespective of whether the sample was taken on the day of intubation or within the next 2 days. Because SP-D has been reported to be associated with death in adults with ARDS, we also examined a model that adjusted for the presence of PARDS and found that SP-D remained independently associated with death (OR = 1.03; CI, 1.01-1.05; P = .001). SP-D was also associated with the duration of mechanical ventilation (Table 4), with length of PICU stay in survivors (Table 5), and with highest OI between days 0 and 2 (Table 6), independent of age, PRISM III, or primary diagnosis.
Figure 3.
Surfactant protein-D levels in survivors and nonsurvivors. Plasma SP-D levels are significantly higher in patients who died (∗P < .001, Wilcoxon Two-Sample test). The number within the bars is the number of patients in each group. See Figures 1 and 2 legends for expansions of abbreviations.
Table 3.
Multivariable Analysis of Association of SP-D With Death in Children With Acute Respiratory Failure
| Variable | OR | 95% CI | P |
|---|---|---|---|
| SP-D | 1.02 | 1.01, 1.04 | .004 |
| Age | 1.08 | 1.00, 1.15 | .040 |
| PRISM III | 1.06 | 1.01, 1.12 | .016 |
| Asthma | 0.73 | 0.16, 3.39 | .686 |
Table 4.
Multivariable Analysis of Association of SP-D With Duration of Mechanical Ventilation in Children With Acute Respiratory Failure
| Variable | Regression Coefficient | 95% CI | P |
|---|---|---|---|
| SP-D | 0.006 | 0.001, 0.011 | .012 |
| Age | 0.007 | −0.008, 0.021 | .390 |
| PRISM III | 0.015 | 0.005, 0.025 | .005 |
| Primary diagnosis | .003 | ||
| Pneumonia | 0.000 | ref | |
| Bronchiolitis | −0.107 | −0.362, 0.148 | .410 |
| Sepsis | 0.094 | −0.122, 0.311 | .394 |
| Asthma | −0.513 | −0.789, −0.238 | < .001 |
| Aspiration pneumonia | −0.255 | −0.582, 0.072 | .127 |
| Other | −0.127 | −0.440, 0.186 | .426 |
Table 5.
Multivariable Analysis of Association of SP-D With PICU Length of Stay in Children With Acute Respiratory Failure
| Variable | Regression Coefficient | 95% CI | P |
|---|---|---|---|
| SP-D | 0.007 | 0.001, 0.012 | .019 |
| Age | 0.004 | −0.011, 0.019 | .598 |
| PRISM III | 0.010 | −0.002, 0.022 | .092 |
| Primary diagnosis | < .001 | ||
| Pneumonia | 0.000 | ref | |
| Bronchiolitis | −0.236 | −0.487, 0.015 | .066 |
| Sepsis | 0.080 | −0.151, 0.312 | .496 |
| Asthma | −0.676 | −0.947, −0.405 | < .001 |
| Aspiration pneumonia | −0.109 | −0.428, 0.210 | .503 |
| Other | −0.213 | −0.542, 0.116 | .204 |
Table 6.
Multivariable Analysis of Association of SP-D With Highest Oxygenation Index in Children With Acute Respiratory Failure
| Variable | Regression Coefficient | 95% CI | P |
|---|---|---|---|
| SP-D | 0.005 | 0.000, 0.009 | .040 |
| Age | 0.011 | −0.005, 0.026 | .176 |
| PRISM III | 0.020 | 0.009, 0.031 | < .001 |
| Primary diagnosis | .047 | ||
| Pneumonia | 0.000 | ref | |
| Bronchiolitis | −0.189 | −0.451, 0.074 | .159 |
| Sepsis | −0.061 | −0.287, 0.165 | .597 |
| Asthma | −0.406 | −0.676, −0.137 | .003 |
| Aspiration pneumonia | −0.276 | −0.613, 0.062 | .110 |
| Other | 0.057 | −0.273, 0.386 | .735 |
Discussion
The severity of alveolar epithelial cell injury is a key determinant of prognosis and severity of ARDS.23 Therefore, studies of measurable plasma biomarkers specific to alveolar epithelial injury are important. SP-D is one of the best characterized markers of alveolar epithelial injury in adults with ARDS, and in the National Heart, Blood, and Lung Institute randomized trial of lung protective ventilation, plasma levels of SP-D declined in the low vs higher tidal volume group.8 The current study of the association of plasma levels of SP-D in children with acute respiratory failure demonstrates a strong association of elevated plasma SP-D with severe PARDS, longer duration of ventilation, and mortality.
This is somewhat different from what has been observed in adults, where SP-D has been reported to be associated with ARDS without any qualification by severity level.10, 11, 12 This discrepancy may be due to a difference between children and adults, between the number of patients with ARDS in those studies and ours (30-40 vs 233), or to the proportion of patients with severe ARDS in the adult studies. At least one of the adult studies has a higher proportion of more severe ARDS patients, because they included only patients with ARDS with a PaO2/Fio2 level < 150.10 The other two papers used the Berlin criteria to identify patients with ARDS but do not report the PaO2/Fio2 levels of their cohorts.11,12
The finding that elevated plasma SP-D levels were associated with increased duration of mechanical ventilation, longer PICU length of stay, and death is similar to what has been reported in adults with acute respiratory failure11 or ARDS,8,24 where increased SP-D has been reported to be associated with death,8,11,24,25 fewer ventilator-free days,8 and weaning within 28 days.11 In adults, SP-D levels have been included in studies using multiple biomarkers to either predict mortality risk24, 25, 26 or identify ARDS subphenotypes.27,28
The day of sample acquisition did not impact the association of SP-D with any of the outcomes examined, suggesting that SP-D is associated with death independent of whether the sample was taken on the day of intubation or within the next 2 days. This finding suggests that samples taken as late as 2 days after intubation might be useful for predicting outcome or identifying subgroups. Unlike inflammatory markers, which quickly decline over time in these patients,16,29 SP-D levels did not decrease across days 0 to 2 (medians and IQRs: 9.4, 5.9-16.9 ng/mL, n = 52; 10.0, 5.7-14.9 ng/mL, n = 174; 14.3, 8.6-22.5, n = 124; respectively), but rather trended upward, with levels on day 2 slightly higher than that on days 0 and 1 (P < .05). Assay of the limited number of day 3 samples (n = 33) indicated that SP-D levels remained elevated (median and IQR: 16.0, 11.2-27.9), suggesting that SP-D may peak several days after intubation (and the initial lung injury) in children with acute respiratory failure. Studies in adults with ARDS also indicate that SP-D levels increase modestly over the first few day,8,10,13 and in one longitudinal study SP-D peaked sometime between days 3 and 7.10 Whether this late peak of plasma SP-D is related to the course of lung injury, which may continue to some degree over time because of ventilator-induced injury, or may be related in some way to the role SP-D plays in the immune response, especially in those patients recovering from acute infection, remains unclear.
The data indicate that SP-D level differs by primary diagnosis; this difference appears to be driven by the lower level of SP-D found in children with a diagnosis of asthma and the higher level in children in the “Other” diagnosis, in whom high levels of SP-D are seen when children are post-BMT or have a primary diagnosis of thoracic trauma, acute chest, or pulmonary edema. Interestingly, unlike data reported for adults,8,30 the level of SP-D does not appear to differ between children with acute respiratory failure that have a primary diagnosis of pneumonia and sepsis. Whether this discrepancy indicates that children’s lungs are less sensitive to damage caused by sepsis than adults, whether there are differences between adults and children in the underlying causes of sepsis, or whether differences in comorbidities between adults and children may result in adult lungs being more sensitive to sepsis-triggered damage are unclear. It will be of interest to determine whether other pediatric cohorts replicate this finding.
Serum SP-D levels increase with age in healthy individuals; however, analyses examining SP-D and age have been done in adults (age range, 18-67) or have compared children (6-9 years of age) with adults.7 In this study, we examined the relationship between SP-D levels and age and found that in critically ill children with acute respiratory failure there is a positive, though relatively weak, correlation between SP-D and age. Examining the relationship of SP-D levels and age is complicated by the fact that even in healthy individuals there is variability between individuals due to genetic as well as environmental factors.7 In the cohort described here, the relationship is also complicated by infection and the degree of lung damage. As expected, when the relationship between SP-D and age was examined in children with a single diagnosis, pneumonia, the correlation between SP-D and age strengthened. These findings suggest that it will be important to consider age when using SP-D as a biomarker in children.
This study has some limitations. First, although this study includes one of the largest cohorts to date of children with acute respiratory failure with or without PARDS with plasma biomarker measurements, our results would have benefited from a larger sample size, particularly when comparing relationships related to the PARDS group as a whole and severe PARDS patients and when using death as the outcome of interest. Also, because our study used data fields obtained as part of the RESTORE study, chest radiograph data are limited to the presence or absence of new bilateral infiltrates. Future studies might benefit from comparing SP-D levels with the degree of lung injury using the recently described more quantitative assessment of pulmonary edema RALE score.31
Conclusions
In summary, elevated SP-D levels are associated with severe PARDS, mortality, OI, duration of mechanical ventilation, and PICU length of stay in children with acute respiratory failure. Future studies should test whether levels of plasma SP-D can be used as a more sensitive indicator of lung injury or response to treatment than the currently available clinical measures and whether it is useful in identifying endotypes in children with PARDS.
Acknowledgments
Author contributions: M. K. D. had full access to all the data in the study and takes responsibility for the integrity of the data and accuracy of the data analysis. M. K. D., H. F., A. S., M. W. Q., and M. A. M. contributed to study concept and design. M. K. D., H. F., A. S., J. K., and M. A. Q. contributed to data collection. H. M. W. and J. K. did the statistical analysis with input from M. K. D. and H. F. All authors were involved with interpretation of data. M. K. D., H. F., J. K., H. M. W., and M. W. Q. were involved with drafting of the manuscript. All authors critically edited the manuscript for important intellectual content. All authors approved the final version before submission.
Financial/nonfinancial disclosures: None declared.
∗BALI and RESTORE Study Investigators and Pediatric Acute Lung Injury and Sepsis Investigators (PALISI) Network Collaborators: We acknowledge the contribution of the subgroup of PALISI members who were BALI study investigators at the sites that participated in the RESTORE study, including: Scot T. Bateman (University of Massachusetts Memorial Children's Medical Center, Worcester, MA), M. D. Berg (University of Arizona Medical Center, Tucson, AZ), Santiago Borasino (Children’s Hospital of Alabama, Birmingham, AL), G. Kris Bysani (Medical City Children's Hospital, Dallas, TX), Allison S. Cowl (Connecticut Children's Medical Center, Hartford, CT), Cindy Darnell Bowens (Children’s Medical Center of Dallas, Dallas, TX), E. Vincent S. Faustino (Yale-New Haven Children’s Hospital, New Haven, CT), Lori D. Fineman (University of California San Francisco Benioff Children’s Hospital at San Francisco, San Francisco, CA), A. J. Godshall (Florida Hospital for Children, Orlando, FL), Ellie Hirshberg (Primary Children’s Medical Center, Salt Lake City, UT), Aileen L. Kirby (Oregon Health & Science University Doernbecher Children's Hospital, Portland, OR), Gwenn E. McLaughlin (Holtz Children’s Hospital, Jackson Health System, Miami, FL), Shivanand Medar (Cohen Children's Medical Center of New York, Hyde Park, NY), Phineas P. Oren (St. Louis Children’s Hospital, St. Louis, MO), James B. Schneider (Cohen Children's Medical Center of New York, Hyde Park, NY), Adam J. Schwarz (Children’s Hospital of Orange County, Orange, CA), Thomas P. Shanley (C. S. Mott Children’s Hospital at the University of Michigan, Ann Arbor, MI), Lauren R. Sorce (Ann & Robert H. Lurie, Children’s Hospital of Chicago, Chicago, IL), Edward J. Truemper (Children’s Hospital and Medical Center, Omaha, NE), Michele A. Vander Heyden (Children's Hospital at Dartmouth, Dartmouth, NH), Kim Wittmayer (Advocate Hope Children’s Hospital, IL), Athena Zuppa (Children’s Hospital of Philadelphia, Philadelphia, PA) and the RESTORE data coordination center led by David Wypij, PhD (Department of Biostatistics, Harvard School of Public Health, Boston, MA; Department of Pediatrics, Harvard Medical School, Boston, MA; Department of Cardiology, Boston Children’s Hospital, Boston, MA).
Role of sponsors: The sponsor had no role in the design, data collection, analyses, interpretation of data, writing of the manuscript, or the decision to submit the study for publication.
Other contributions: We thank the study participants and their families and guardians for their participation in this study.
Additional information: The e-Appendix, e-Figures, and e-Tables can be found in the Supplemental Materials section of the online article.
Footnotes
FUNDING/SUPPORT: This study was funded by a grant from the NIH awarded to Drs. Dahmer, Flori and Quasney (R01HL095410). The parent study was supported by grants from the NIH awarded to Drs. Curley and Wypij (U01HL086622, U01 HL086649).
Contributor Information
Michael W. Quasney, Email: mquasney@med.umich.edu.
BALI and RESTORE Study Investigators and Pediatric Acute Lung Injury and Sepsis Investigators (PALISI) Network:
Scot T. Bateman, M.D. Berg, Santiago Borasino, G. Kris Bysani, Allison S. Cowl, Cindy Darnell Bowens, E. Vincent S. Faustino, Lori D. Fineman, A.J. Godshall, Ellie Hirshberg, Aileen L. Kirby, Gwenn E. McLaughlin, Shivanand Medar, Phineas P. Oren, James B. Schneider, Adam J. Schwarz, Thomas P. Shanley, Lauren R. Sorce, Edward J. Truemper, Michele A. Vander Heyden, Kim Wittmayer, Athena Zuppa, and David Wypij
Supplementary Data
References
- 1.Crouch E.C. Structure, biologic properties, and expression of surfactant protein D (SP-D) Biochim Biophys Acta. 1998;1408(2-3):278–289. doi: 10.1016/s0925-4439(98)00073-8. [DOI] [PubMed] [Google Scholar]
- 2.Crouch E., Wright J.R. Surfactant proteins a and d and pulmonary host defense. Annu Rev Physiol. 2001;63:521–554. doi: 10.1146/annurev.physiol.63.1.521. [DOI] [PubMed] [Google Scholar]
- 3.Foo S.S., Reading P.C., Jaillon S., Mantovani A., Mahalingam S. Pentraxins and collectins: friend or foe during pathogen invasion? Trends Microbiol. 2015;23(12):799–811. doi: 10.1016/j.tim.2015.09.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Wright J.R. Immunoregulatory functions of surfactant proteins. Nat Rev Immunol. 2005;5(1):58–68. doi: 10.1038/nri1528. [DOI] [PubMed] [Google Scholar]
- 5.Sano H., Kuroki Y. The lung collectins, SP-A and SP-D, modulate pulmonary innate immunity. Mol Immunol. 2005;42(3):279–287. doi: 10.1016/j.molimm.2004.07.014. [DOI] [PubMed] [Google Scholar]
- 6.Sorensen G.L. Surfactant protein D in respiratory and non-respiratory diseases. Front Med (Lausanne) 2018;5:18. doi: 10.3389/fmed.2018.00018. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Sorensen G.L., Hjelmborg J.B., Kyvik K.O. Genetic and environmental influences of surfactant protein D serum levels. Am J Physiol Lung Cell Mol Physiol. 2006;290(5):L1010–1017. doi: 10.1152/ajplung.00487.2005. [DOI] [PubMed] [Google Scholar]
- 8.Eisner M.D., Parsons P., Matthay M.A., Ware L., Greene K. Plasma surfactant protein levels and clinical outcomes in patients with acute lung injury. Thorax. 2003;58(11):983–988. doi: 10.1136/thorax.58.11.983. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Sorensen G.L., Husby S., Holmskov U. Surfactant protein A and surfactant protein D variation in pulmonary disease. Immunobiology. 2007;212(4-5):381–416. doi: 10.1016/j.imbio.2007.01.003. [DOI] [PubMed] [Google Scholar]
- 10.Greene K.E., Wright J.R., Steinberg K.P. Serial changes in surfactant-associated proteins in lung and serum before and after onset of ARDS. Am J Respir Crit Care Med. 1999;160(6):1843–1850. doi: 10.1164/ajrccm.160.6.9901117. [DOI] [PubMed] [Google Scholar]
- 11.Jensen J.S., Itenov T.S., Thormar K.M. Prediction of non-recovery from ventilator-demanding acute respiratory failure, ARDS and death using lung damage biomarkers: data from a 1200-patient critical care randomized trial. Ann Intensive Care. 2016;6(1):114. doi: 10.1186/s13613-016-0212-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Park J., Pabon M., Choi A.M.K. Plasma surfactant protein-D as a diagnostic biomarker for acute respiratory distress syndrome: validation in US and Korean cohorts. BMC Pulm Med. 2017;17(1):204. doi: 10.1186/s12890-017-0532-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Determann R.M., Royakkers A.A., Haitsma J.J. Plasma levels of surfactant protein D and KL-6 for evaluation of lung injury in critically ill mechanically ventilated patients. BMC Pulm Med. 2010;10:6. doi: 10.1186/1471-2466-10-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Dahl M., Holmskov U., Husby S., Juvonen P.O. Surfactant protein D levels in umbilical cord blood and capillary blood of premature infants: the influence of perinatal factors. Pediatr Res. 2006;59(6):806–810. doi: 10.1203/01.pdr.0000219122.81734.03. [DOI] [PubMed] [Google Scholar]
- 15.Todd D.A., Marsh M.J., George A. Surfactant phospholipids, surfactant proteins, and inflammatory markers during acute lung injury in children. Pediatr Crit Care Med. 2010;11(1):82–91. doi: 10.1097/PCC.0b013e3181ae5a4c. [DOI] [PubMed] [Google Scholar]
- 16.Dahmer M.K., Quasney M.W., Sapru A. Interleukin-1 receptor antagonist is associated with pediatric acute respiratory distress syndrome and worse outcomes in children with acute respiratory failure. Pediatr Crit Care Med. 2018;19:930–938. doi: 10.1097/PCC.0000000000001680. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Curley M.A., Wypij D., Watson R.S. Protocolized sedation vs usual care in pediatric patients mechanically ventilated for acute respiratory failure: a randomized clinical trial. JAMA. 2015;313(4):379–389. doi: 10.1001/jama.2014.18399. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Nagae H., Takahashi H., Kuroki Y. Enzyme-linked immunosorbent assay using F(ab')2 fragment for the detection of human pulmonary surfactant protein D in sera. Clin Chim Acta. 1997;266(2):157–171. doi: 10.1016/s0009-8981(97)00124-1. [DOI] [PubMed] [Google Scholar]
- 19.Pediatric Acute Lung ICCG Pediatric acute respiratory distress syndrome: consensus recommendations from th epediatric acute lung injury consensus conference. Pediatr Crit Care Med. 2015;16:428–439. doi: 10.1097/PCC.0000000000000350. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Bernard G.R., Artigas A., Brigham K.L. The American-European Consensus Conference on ARDS. Definitions, mechanisms, relevant outcomes, and clinical trial coordination. Am J Respir Crit Care Med. 1994;149(3 Pt 1):818–824. doi: 10.1164/ajrccm.149.3.7509706. [DOI] [PubMed] [Google Scholar]
- 21.Khemani R.G., Thomas N.J., Venkatachalam V. Comparison of SpO2 to PaO2 based markers of lung disease severity for children with acute lung injury. Crit Care Med. 2012;40(4):1309–1316. doi: 10.1097/CCM.0b013e31823bc61b. [DOI] [PubMed] [Google Scholar]
- 22.Elliott A.C., Hynan L.S. A SAS((R)) macro implementation of a multiple comparison post hoc test for a Kruskal-Wallis analysis. Comput Methods Programs Biomed. 2011;102(1):75–80. doi: 10.1016/j.cmpb.2010.11.002. [DOI] [PubMed] [Google Scholar]
- 23.Matthay M.A., Zemans R.L., Zimmerman G.A. Acute respiratory distress syndrome. Nat Rev Dis Primers. 2019;5(1):18. doi: 10.1038/s41572-019-0069-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Ware L.B., Koyama T., Billheimer D.D. Prognostic and pathogenetic value of combining clinical and biochemical indices in patients with acute lung injury. Chest. 2010;137(2):288–296. doi: 10.1378/chest.09-1484. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Calfee C.S., Ware L.B., Glidden D.V. Use of risk reclassification with multiple biomarkers improves mortality prediction in acute lung injury. Crit Care Med. 2011;39(4):711–717. doi: 10.1097/CCM.0b013e318207ec3c. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Zhao Z., Wickersham N., Kangelaris K.N. External validation of a biomarker and clinical prediction model for hospital mortality in acute respiratory distress syndrome. Intensive Care Med. 2017;43(8):1123–1131. doi: 10.1007/s00134-017-4854-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Calfee C.S., Delucchi K., Parsons P.E. Subphenotypes in acute respiratory distress syndrome: latent class analysis of data from two randomised controlled trials. Lancet Respir Med. 2014;2(8):611–620. doi: 10.1016/S2213-2600(14)70097-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Famous K.R., Delucchi K., Ware L.B. Acute respiratory distress syndrome subphenotypes respond differently to randomized fluid management strategy. Am J Respir Crit Care Med. 2017;195(3):331–338. doi: 10.1164/rccm.201603-0645OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Flori H., Sapru A., Quasney M.W. A prospective investigation of interleukin-8 levels in pediatric acute respiratory failure and acute respiratory distress syndrome. Crit Care. 2019;23(1):128. doi: 10.1186/s13054-019-2342-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Calfee C.S., Janz D.R., Bernard G.R. Distinct molecular phenotypes of direct vs indirect ARDS in single-center and multicenter studies. Chest. 2015;147(6):1539–1548. doi: 10.1378/chest.14-2454. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Warren M.A., Zhao Z., Koyama T. Severity scoring of lung oedema on the chest radiograph is associated with clinical outcomes in ARDS. Thorax. 2018;73(9):840–846. doi: 10.1136/thoraxjnl-2017-211280. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.