Abstract WP222: Temporally Differentially Expressed MikroRNAs In Serum Predict Recovery In Different Acute Brain Injuries

Background: Brain plasticity is important processes in recovering after different types of acute brain injuries such as aneurysmal subarachnoid hemorrhage (aSAH), ischemic stroke (IS) and traumatic brain injury (TBI). Knowledge gaps still exist which miRNAs contribute to recovery of the acute brain injuries. Hypotheses: Temporally differentially expressed (DE) miRNAs after acute brain injuries may have association to outcome irrespective the type of the acute brain injury. MiRNAs that are DE across different type of brain injuries may reveal important conserved associations that can reflect important biological mechanisms and serve as a biomarker. Methods: Prospective cohort (n=24) consisted of IS (n=8), aSAH (n=8) and TBI (n=8) patients. Two serum samples were collected per patient (early 24-48h and late 120-192h after injury). Outcome was measured 90 days after injury (mRS favorable 0-3 (n=15), unfavorable 4-6 (n=9). MiRNAs were extracted from the samples (total n = 48 samples). Sequencing was performed and DESeq2 was used to identify DE miRNAs. The miRNA putative target genes were predicted with miRWalk. Normalized expression values of identified DE miRNAs were used and linear canonical discriminant analysis (LDA) was performed. Canonical scores were used to build combinatory biomarker with logistic modelling predicting outcome. Results: We identified 22 temporally DE miRNAs (p<0.05) that were in common across all acute brain injury types when compared between favorable and unfavorable groups. From this pool miRWalk target analysis identified 4 miRNAs (hsa-miR-146b-3p, hsa-miR-485-3p, hsa-miR-5010-5p, hsa-miR-485-5p) that targeted to known plasticity mechanism. LDA of these four miRNAs resulted an equation with canonical scores: 0.636 x [hsa-miR-146b-3p] + 0.576 x [hsa-miR-485-3p] + 0.652 x [hsa-miR-5010-5p] + 0.372 x [hsa-miR-485-5p]. The receiver operating characteristic curve generated showed AUC = 94.1%, 95% CI = (0.849, 1.00), p = 0.016 Conclusions: Results show that combinatory biomarker of identified miRNAs perform well across different type of brain injuries. Validation in larger cohort is justified. Identified common miRNAs also provide new targets for mechanistical validation in disease models of TBI and stroke.