SAN DIEGO — Endotypes derived from individualized gene expression patterns of critically ill pediatric patients showed that certain biomarkers were linked with mortality, results from a late-breaking trial found.
In a sample of 67 patients 10 years and under with acute hypoxemic respiratory failure, those assigned to endotype A (characterized by repression of genes associated with adaptive immunity and glucocorticoid receptor signaling) had higher 28-day mortalities than the endotype B group (odds ratio 8.0, 95% CI 1.6-41.0), reported Nadir Yehya, MD, of the Children’s Hospital of Philadelphia, and colleagues.
Additionally, the endotype A group had nearly threefold the risk of having a complicated disease course compared with the endotype B group (OR 4.2, 95% CI 1.2-14.9), the team reported here at the Society of Critical Care Medicine (SCCM) meeting and online in Pediatric Critical Care Medicine.
And these associations persisted when adjusting for age, severity of illness, and partial pressure arterial oxygen and fraction of inspired oxygen (PaO2/FIO2), Yehya noted in his presentation. “Biomarker-based endotypes in adult patients with acute respiratory distress syndrome (ARDS) have been shown to reduce heterogeneity, identify meaningful subgroups, and — probably most importantly — to identify underlying biology with differential response to therapy.”
Prior research has demonstrated that adult patients with ARDS and biomarker-defined endotypes respond differently to positive end-expiratory pressure and fluid management, although less research has been conducted in pediatric populations regarding these patterns, he said.
Deanna Behrens, MD, of Advocate Children’s Medical Group in Chicago, who was not involved with the study, said that although Yehya and colleagues didn’t examine which treatments might be more effective in children with either endotype, the results could inform future trials that did so.
“I would look forward to seeing what they’re doing with this in the future, particularly as our genomic testing becomes more sophisticated,” Behrens told MedPage Today. “It might be that we’ll be able to have some of these tests on hand more readily, and in the case that it helps us identify sicker kids, there would be potential interventions we could do.”
In a previous study, Yehya and colleagues differentiated endotypes A and B in pediatric septic shock patients using peripheral messenger RNA expression. In that study, endotype A was associated with poorer outcomes compared with endotype B, he said. The new study is a re-analysis of this array data examining children with acute hypoxemic respiratory failure.
The researchers used Pediatric Risk for Mortality (PRISM) III and Pediatric Sepsis Biomarker Risk Model (PERSEVERE) scores to determine general and septic shock-related mortality risk from the day of admission to a pediatric intensive care unit (PICU). Organ failure (identified with Goldstein criteria) was tracked for 1 week after admission and mortality was measured for 28 days. Blood was obtained within 24 hours of admission, and subsequent clinical and laboratory data were collected daily, the authors noted.
The data was collected from the National Center for Biotechnology Information Gene Expression Omnibus of children ≤10 years who were admitted to the PICU meeting 2005 septic shock criteria. Of this sample, patients were considered to have acute hypoxemic respiratory failure with PaO2/FIO2 ≤200 in the first week, which would be considered moderate to severe ARDS, Yehya said.
In total, 23 patients were assigned to endotype A and 44 to endotype B. Patients in the former group were younger on average (0.8 years vs 4.3 years) and more often male (64% vs 52%), and the endotype A group also had higher PRISM III (22 vs 15) and PERSEVERE scores (0.227 vs 0.101) compared with endotype B, the researchers reported.
The authors noted it may be that these endotypes were more indicative with greater severity of illness. Yehya added that it could also be that patients with sepsis and acute hypoxemic respiratory failure have shared underlying biology that may predict mortality risk.
The researchers said the study was limited because it was observational in nature and no protocol of care was established for this cohort of patients. Additionally, chest x-rays were not available, so the team had to rely on Berlin criteria to determine whether patients had ARDS. In addition, detailed ventilator data and oxygenation indexes were not available to the researchers either, they noted.
This study and the authors were supported in part by the National Institutes of Health.
Co-authors received support from Genefluidics, Therabron, and CareFusion.