Receptor for advanced glycation end-products and ARDS prediction: a multicentre observational study

Abstract

Acute respiratory distress syndrome (ARDS) prediction remains challenging despite available clinical scores. To assess soluble receptor for advanced glycation end-products (sRAGE), a marker of lung epithelial injury, as a predictor of ARDS in a high-risk population, adult patients with at least one ARDS risk factor upon admission to participating intensive care units (ICUs) were enrolled in a multicentre, prospective study between June 2014 and January 2015. Plasma sRAGE and endogenous secretory RAGE (esRAGE) were measured at baseline (ICU admission) and 24 hours later (day one). Four AGER candidate single nucleotide polymorphisms (SNPs) were also assayed because of previous reports of functionality (rs1800625, rs1800624, rs3134940, and rs2070600). The primary outcome was ARDS development within seven days. Of 500 patients enrolled, 464 patients were analysed, and 59 developed ARDS by day seven. Higher baseline and day one plasma sRAGE, but not esRAGE, were independently associated with increased ARDS risk. AGER SNP rs2070600 (Ser/Ser) was associated with increased ARDS risk and higher plasma sRAGE in this cohort, although confirmatory studies are needed to assess the role of AGER SNPs in ARDS prediction. These findings suggest that among at-risk ICU patients, higher plasma sRAGE may identify those who are more likely to develop ARDS.

Figure 1

Flow diagram of the study. Personnel shortage was another identified reason for non-enrollment of eligible subjects, and some subjects were missed without a given explanation.

Figure 2

Plasma biomarker levels according to ARDS development. Patients who developed ARDS at least 24 hours after the first sample draw (n = 59) had statistically significant increased (A) baseline plasma sRAGE, (B) plasma sRAGE on day one, and (C) day one-to-day zero plasma sRAGE ratio than those who did not develop ARDS (n = 405). (D) Baseline plasma esRAGE, (E) plasma esRAGE on day one, and (F) day one-to-day zero plasma esRAGE ratio were similar between groups.

Figure 3

Receiver operating characteristic (ROC) curves. (A) Baseline plasma sRAGE (area under ROC curve (AUROC), 0.74; 95% CI, [0.68–0.80] and (B) plasma sRAGE on day one (AUROC, 0.82 [0.76–0.88]) each showed good discrimination between those who developed ARDS and those who did not, but (C) the LIPS was poorly discriminative in this population of patients at risk of developing ARDS (AUROC, 0.57 [0.49–0.65]).

Figure 4

The cumulative proportion of patients who did not develop ARDS within seven days of admission to the ICU for (A) patients with baseline plasma sRAGE above or below 1,033 pg/mL (the median value of baseline plasma sRAGE in patients from our cohort) and (B) patients with or without homozygous SNP rs2070600 (the Ser/Ser genotype) within the gene coding for RAGE.