SCCM Resource Library-Biomarker Trajectories of Direct and Indirect Acute Respiratory Distress Syndrome in Children

Video Transcription

Hi, everybody. Today, we're going to be talking about biomarker trajectories of direct and indirect acute respiratory distress syndrome in children. My name is Nader Yahya. These  are my disclosures. For those of you who track this sort of thing, this talk is fully in the spirit of the immune system is at it again, with a dash of, hey, at least we showed that  this method can produce results. That's not nothing. So in a surprise to nobody listening to this talk, the heterogeneity of ARDS has hampered  clinical trials. And attempts at subphenotyping ARDS is an attempt to mitigate this heterogeneity.  People have tried clinical subtyping, such as direct versus indirect, which is fairly nonspecific and runs a pretty high risk of misclassification. Imaging subtypes, such  as focal and non-focal, which were operationalized in the live trial, also risk misclassification.  20% of patients in live, for example, were misclassified. And then reanalyses of several  ARDS network trials, as well as other cohorts, have demonstrated biomarker-based subphenotypes,  grossly labeled hypo and hyperinflammatory. There's a lot of appeal to biomarker-based subtyping. Biomarkers are felt to be objective and reproducible. There's also the hint of  mechanism, which is available in the biomarkers. The disadvantages of biomarkers is a different  type of nonspecificity. What exactly are we measuring? The most informative biomarkers, for example, in the hypo and hyperinflammatory subtypes are those which represent shock and  innate immunity. For example, IL-6, IL-8, and TNFR-1, which are elevated in the hyperinflammatory  subphenotype, are really just reflections of innate immunity. And the systolic blood pressure,  protein C, and bicarbonate are reflections of degree of shock. So neither of these characteristics  are particularly ARDS-specific. We also don't really have a good handle on what cutoffs for  any individual biomarker we should be using if we want to enrich for a subphenotype. And finally,  we don't really have a good sense of whether we should be using single biomarkers, or panels of  biomarkers, or measuring at which time point. Tissue-based biomarkers have received a little  bit of attention in this subphenotyping strategy. People have focused on endothelial biomarkers,  such as angiopoietin-2 and von Willebrand factor. The soluble receptor for advanced glycation end  products, or SRAGE, has received attention as a potential marker of type 1 alveolar epithelial  damage. And surfactant protein D is an established marker of type 2 alveolar epithelial damage.  Of those three, the origins of SRAGE and ARDS are probably the most controversial. It's definitely  expressed in type 1 epithelia, and it's inversely correlated with alveolar fluid clearance. However,  it's also definitely expressed in a ton of other places, such as the endothelium, and in several cohort studies does not discriminate direct ARDS. Subtyping is  interesting, but can tissue-specific biomarkers perform subtyping better? And finally, all of  these studies have been conducted in adults, and so we need to establish the relevance of  tissue-specific biomarkers in pediatric ARDS. Therefore, we aimed in our study to compare the longitudinal trajectory of three proteins reflecting tissue damage in pediatric ARDS.  We used angiopoietin 2 as a confirmed marker of endothelial damage. We used SRAGE as a putative  marker of type 1 epithelial damage, and we used surfactant protein D as a confirmed marker of type 2 epithelial damage. We also wanted to investigate whether or not we could group  subjects into subtypes which shared longitudinal biomarker profiles, mostly just to assess the feasibility of this approach to see whether we could generate some construct validity,  and to compare the clinical characteristics of the cohorts that we found. This was a  prospective cohort study with intubated children with Berlin-defined ARDS, with 279 subjects with  blood sampling on day 0, down to 207 available for blood sampling on day 7. Between day 0 and  day 7, the drop-off was about evenly split between patients who rapidly improved as well as patients  who died within seven days. We examined the differences of biomarker levels between direct and indirect ARDS, between Berlin severity categories, the number of organ failures,  and mortality. We performed univariate analysis comparing those characteristics, as well as multivariate analysis in which we log-transformed and adjusted for known  confounders between the association of biomarker levels and mortality. We finally performed latent  class growth analysis in order to identify longitudinal biomarker subtypes. When we  compared direct versus indirect ARDS, surfactant protein D was higher in direct ARDS, particularly  on day 0 and day 3. There was no difference by day 7. SRAGE and angiopoietin 2 were consistently  higher in indirect ARDS. Surfactant protein D and SRAGE also tracked with Berlin severity  categories, with increasing biomarker levels for SPD and SRAGE across increasing Berlin severity  categories. SRAGE and angiopoietin 2 also strongly associated with increasing number of non-pulmonary  organ failures, whereas SPD showed no association with non-pulmonary organ failures. All of the  biomarkers were elevated in non-survivors. SPD showed elevation in non-survivor on day 0, 3,  and 7. SRAGE and angiopoietin 2 were strongly higher in non-survivors at all three time points  as well. After adjusting for confounders and looked at in a different way, all three biomarkers were  again higher in survivors, although the association between SPD and survival status was not significant  after adjustment for confounders. SRAGE and angiopoietin 2 retained significance after adjusting for confounders. Latent class growth analysis demonstrated three subphenotypes  identified based on longitudinal biomarker trajectory. Group 1 in the first column,  using that as a reference, comprised the majority of the cohort. Group 2, when compared against group 1, showed higher levels of angiopoietin 2. And group 3, which  had the worst outcome, had higher levels of surfactant protein D and the highest initial levels of SRAGE. Overall, surfactant protein D behaved very similar to what you would expect  as a marker of epithelial damage. It was higher with worse oxygenation, it was higher in direct ARDS, and it had a decent association with mortality, particularly in direct ARDS.  Angiopoietin 2 behaved very much as you would expect for a marker of endothelial damage, where it has minimal to no correlation with oxygenation. It had a very strong correlation  with multisystem organ failure. It had higher levels in indirect ARDS and a strong association  with mortality. SRAGE behaved as if it had components of both, where it had an association with both oxygenation as well as multisystem organ failure. It was higher in indirect,  more similar to angiopoietin 2 rather than surfactant protein D, and it also showed a strong  association with mortality. Overall, from our study, we conclude that surfactant protein D is higher in direct ARDS and that SRAGE and angiopoietin 2 are higher in indirect ARDS.  All three predicted mortality to varying degrees, although the ones associated with organ failure  did a better job at predicting mortality than did surfactant protein D. Finally, we showed that three subtypes could be identified using longitudinal biomarker trajectories rather  than relying on biomarkers at a single time point, and these subphenotypes demonstrated clinical differences that deserve further exploration as a way to subcategorize ARDS.

Video Summary

In this video, the speaker discusses biomarker trajectories in pediatric acute respiratory distress syndrome (ARDS). They explain that subtyping ARDS is challenging due to its heterogeneity, and clinical and imaging subtyping methods have limitations. Biomarker-based subtyping is appealing but also has challenges, such as nonspecificity and the lack of defined cutoffs. The speaker explores tissue-specific biomarkers, including angiopoietin 2, SRAGE, and surfactant protein D, and their potential for subtyping ARDS. The study compares the longitudinal trajectory of these biomarkers in pediatric ARDS, identifies subtypes based on biomarker profiles, and examines their association with clinical characteristics and mortality. The findings suggest that different biomarkers are associated with specific ARDS subtypes and predict mortality to varying degrees.

Asset Subtitle

Pulmonary, Pediatrics, 2022

Asset Caption

INTRODUCTION: Acute respiratory distress syndrome (ARDS) can be classified according to direct (pulmonary) or indirect (non-pulmonary) etiology in order to mitigate heterogeneity. Surfactant protein D (SP-D) and angiopoietin-2 (Ang-2) are proposed indicators of direct and indirect ARDS, respectively, whereas utility of the soluble receptor for advanced glycation end-products (sRAGE) is controversial. The majority of studies have been conducted in adults, and differences in the biomarker levels between direct and indirect ARDS have not been investigated in children. Therefore, we aimed to compare the longitudinal biomarker profiles of direct and indirect pediatric ARDS using SP-D, sRAGE, and Ang-2. METHODS: This was a prospective cohort study of intubated children with ARDS, with biomarker measurements on day 0 (n = 279), day 3 (n = 266), and day 7 (n = 207). We assessed the relationship between biomarkers and direct/indirect ARDS classification, ARDS severity, non-pulmonary organ failures, and PICU mortality using Fisher’s exact test or Cuzick’s non-parametric test of trend, as appropriate. Based on values for sRAGE from a prior study, we estimated that 275 subjects would provide 80% power to detect a difference of 1 ng/mL in biomarker levels between direct and indirect ARDS at α = 0.017. RESULTS: There were 279 subjects with ARDS, with 64 (22.9%) non-survivors. SP-D was higher in direct ARDS on day 0 (p < 0.001) and was strongly associated with mortality in direct ARDS, with the highest SP-D levels in subjects who died of refractory hypoxemia. sRAGE was higher in indirect ARDS on days 3 and 7 (both p < 0.001) and correlated with both ARDS severity and with non-pulmonary organ failures. Ang-2 was higher in indirect ARDS at all timepoints (all p < 0.001) and correlated with non-pulmonary organ failures. sRAGE and Ang-2 were associated with mortality in both direct and indirect ARDS. Conclusions were unchanged in sensitivity analyses using only data from the 207 subjects with measurements at all three timepoints. CONCLUSIONS: SP-D and Ang-2 performed consistent with their proposed roles as biomarkers of direct and indirect pediatric ARDS, respectively. sRAGE was higher in indirect ARDS, correlated most strongly with multisystem organ failure, and should not be considered a marker of direct ARDS.

Meta Tag

Content Type Presentation

Knowledge Area Pulmonary

Knowledge Area Pediatrics

Knowledge Level Advanced

Membership Level Select

Tag Acute Respiratory Distress Syndrome ARDS

Tag Pediatrics

Tag Acute Lung Injury

Tag Mortality

Tag Respiratory Failure

Year 2022

Keywords

biomarker trajectories

pediatric acute respiratory distress syndrome

subtyping ARDS

tissue-specific biomarkers

mortality prediction

	Society of Critical Care Medicine 500 Midway Drive Mount Prospect, IL 60056 USA Phone: +1 847 827-6888 Fax: +1 847 439-7226 Email: support@sccm.org	Contact Us About SCCM Newsroom Advertising & Sponsorship DONATE	MySCCM LearnICU Patients & Families Surviving Sepsis Campaign Critical Care Societies Collaborative	GET OUR NEWSLETTER
© Society of Critical Care Medicine. All rights reserved. \| Privacy Statement \| Terms & Conditions The Society of Critical Care Medicine, SCCM, and Critical Care Congress are registered trademarks of the Society of Critical Care Medicine.