[Skip to Content]
Access to paid content on this site is currently suspended due to excessive activity being detected from your IP address 34.236.145.124. Please contact the publisher to request reinstatement.
[Skip to Content Landing]
Figure 1.
Levels of MicroRNA 124-3p (miR-124-3p) According to Targeted Temperature and Neurologic Outcome
Levels of MicroRNA 124-3p (miR-124-3p) According to Targeted Temperature and Neurologic Outcome

Circulating levels of miR-124-3p were assessed 48 hours after the return of spontaneous circulation using quantitative polymerase chain reaction in 529 patients of the validation cohort. Comparisons of miR-124-3p levels between patients in the 33°C or 36°C treatment groups, between all patients with good and poor outcomes and between patients in each temperature group with good and poor outcomes are shown. Good outcomes indicate a cerebral performance category (CPC) score of 1 (no neurologic sequelae) or 2 (moderate neurologic sequelae); poor outcomes, a CPC score of 3 (severe neurologic sequelae), 4 (coma), or 5 (death). The miR-124-3p levels are expressed in a log scale. The lower boundary of the boxes indicate the 25th percentile; horizontal line within the box, median; higher boundary of the box, 75th percentile; error bars, 90th and 10th percentiles; and individual points, outliers.

Figure 2.
Association of Neurologic Outcome and Death With miR-124-3p Level
Association of Neurologic Outcome and Death With miR-124-3p Level

Circulating levels of miR-124-3p were assessed using quantitative polymerase chain reaction in blood samples obtained 48 hours after the return of spontaneous circulation in 529 patients of the validation cohort. A-C, The prognostic value for neurologic outcome at 6 months assessed by dichotomized cerebral performance category score was determined using logistic regression for all patients and those in the 33°C (n = 231) and 36°C (n = 298) treatment groups. Areas under the receiver operating characteristics curve (95% CI) are shown for each group. D, The prognostic value of microRNA 124-3p (miR-124-3p) levels for death at the end of the trial was determined using Kaplan-Meier curves and the log-rank test. Quartile 1 indicates patients with miR-124-3p serum levels of 3 to 688 copies/μL; quartile 2, 690 to 1565 copies/μL; quartile 3, 1572 to 4817 copies/μL; and quartile 4, 4882 to 83 373 copies/μL.

Table 1.  
Clinical Features of 579 Patients With Good and Poor Neurologic Outcomes Within the miRNA Cohort
Clinical Features of 579 Patients With Good and Poor Neurologic Outcomes Within the miRNA Cohort
Table 2.  
Prognostic Value for Study Outcomes
Prognostic Value for Study Outcomes
1.
Lemiale  V, Dumas  F, Mongardon  N,  et al.  Intensive care unit mortality after cardiac arrest: the relative contribution of shock and brain injury in a large cohort.  Intensive Care Med. 2013;39(11):1972-1980.PubMedGoogle ScholarCrossref
2.
Dragancea  I, Rundgren  M, Englund  E, Friberg  H, Cronberg  T.  The influence of induced hypothermia and delayed prognostication on the mode of death after cardiac arrest.  Resuscitation. 2013;84(3):337-342.PubMedGoogle ScholarCrossref
3.
Laver  S, Farrow  C, Turner  D, Nolan  J.  Mode of death after admission to an intensive care unit following cardiac arrest.  Intensive Care Med. 2004;30(11):2126-2128.PubMedGoogle ScholarCrossref
4.
Sandroni  C, Cavallaro  F, Callaway  CW,  et al.  Predictors of poor neurological outcome in adult comatose survivors of cardiac arrest: a systematic review and meta-analysis, part 2: patients treated with therapeutic hypothermia.  Resuscitation. 2013;84(10):1324-1338.PubMedGoogle ScholarCrossref
5.
Friberg  H, Cronberg  T, Dünser  MW, Duranteau  J, Horn  J, Oddo  M.  Survey on current practices for neurological prognostication after cardiac arrest.  Resuscitation. 2015;90:158-162.PubMedGoogle ScholarCrossref
6.
Sandroni  C, Cariou  A, Cavallaro  F,  et al.  Prognostication in comatose survivors of cardiac arrest: an advisory statement from the European Resuscitation Council and the European Society of Intensive Care Medicine.  Resuscitation. 2014;85(12):1779-1789.PubMedGoogle ScholarCrossref
7.
Nolan  JP, Soar  J, Cariou  A,  et al.  European Resuscitation Council and European Society of Intensive Care Medicine Guidelines for Post-resuscitation Care 2015: section 5 of the European Resuscitation Council Resuscitation Guidelines 2015.  Resuscitation. 2015.PubMedGoogle Scholar
8.
Stammet  P, Collignon  O, Hassager  C,  et al; TTM-Trial Investigators.  Neuron-specific enolase as a predictor of death or poor neurological outcome after out-of-hospital cardiac arrest and targeted temperature management at 33°C and 36°C.  J Am Coll Cardiol. 2015;65(19):2104-2114.PubMedGoogle ScholarCrossref
9.
Devaux  Y, Stammet  P, Friberg  H,  et al; Biomarker Subcommittee of TTM trial (Target Temperature Management After Cardiac Arrest, NCT01020916).  MicroRNAs: new biomarkers and therapeutic targets after cardiac arrest?  Crit Care. 2015;19(1):54.PubMedGoogle ScholarCrossref
10.
Mitchell  PS, Parkin  RK, Kroh  EM,  et al.  Circulating microRNAs as stable blood-based markers for cancer detection.  Proc Natl Acad Sci U S A. 2008;105(30):10513-10518.PubMedGoogle ScholarCrossref
11.
Goretti  E, Wagner  DR, Devaux  Y.  miRNAs as biomarkers of myocardial infarction: a step forward towards personalized medicine?  Trends Mol Med. 2014;20(12):716-725.PubMedGoogle ScholarCrossref
12.
Stammet  P, Goretti  E, Vausort  M, Zhang  L, Wagner  DR, Devaux  Y.  Circulating microRNAs after cardiac arrest.  Crit Care Med. 2012;40(12):3209-3214.PubMedGoogle ScholarCrossref
13.
Gilje  P, Gidlöf  O, Rundgren  M,  et al.  The brain-enriched microRNA miR-124 in plasma predicts neurological outcome after cardiac arrest.  Crit Care. 2014;18(2):R40.PubMedGoogle ScholarCrossref
14.
ClinicalTrials.gov. Target Temperature Management After Cardiac Arrest. NCT01020916. https://clinicaltrials.gov/ct2/show/NCT01020916. Accessed April 5, 2016.
15.
Nielsen  N, Wetterslev  J, Cronberg  T,  et al; TTM Trial Investigators.  Targeted temperature management at 33°C vs 36°C after cardiac arrest.  N Engl J Med. 2013;369(23):2197-2206.PubMedGoogle ScholarCrossref
16.
Nielsen  N, Wetterslev  J, al-Subaie  N,  et al.  Target temperature management after out-of-hospital cardiac arrest: a randomized, parallel-group, assessor-blinded clinical trial—rationale and design.  Am Heart J. 2012;163(4):541-548.PubMedGoogle ScholarCrossref
17.
Nielsen  N, Winkel  P, Cronberg  T,  et al.  Detailed statistical analysis plan for the target temperature management after out-of-hospital cardiac arrest trial.  Trials. 2013;14(1):300.PubMedGoogle ScholarCrossref
18.
World Medical Association.  World Medical Association Declaration of Helsinki: ethical principles for medical research involving human subjects.  JAMA. 2013;310(20):2191-2194.PubMedGoogle ScholarCrossref
19.
Booth  CM, Boone  RH, Tomlinson  G, Detsky  AS.  Is this patient dead, vegetative, or severely neurologically impaired? assessing outcome for comatose survivors of cardiac arrest.  JAMA. 2004;291(7):870-879.PubMedGoogle ScholarCrossref
20.
Kozomara  A, Griffiths-Jones  S.  miRBase: integrating microRNA annotation and deep-sequencing data.  Nucleic Acids Res. 2011;39(Database issue):D152-D157.PubMedGoogle ScholarCrossref
21.
Robinson  MD, McCarthy  DJ, Smyth  GK.  edgeR: a Bioconductor package for differential expression analysis of digital gene expression data.  Bioinformatics. 2010;26(1):139-140.PubMedGoogle ScholarCrossref
22.
Robinson  MD, Oshlack  A.  A scaling normalization method for differential expression analysis of RNA-seq data.  Genome Biol. 2010;11(3):R25.PubMedGoogle ScholarCrossref
23.
Devaux  Y, Vausort  M, Goretti  E,  et al.  Use of circulating microRNAs to diagnose acute myocardial infarction.  Clin Chem. 2012;58(3):559-567.PubMedGoogle ScholarCrossref
24.
Harrell  FE.  Regression Modeling Strategies. New York, NY: Springer Science & Business Media; 2001.
25.
Stammet  P, Wagner  DR, Gilson  G, Devaux  Y.  Modeling serum level of s100β and bispectral index to predict outcome after cardiac arrest.  J Am Coll Cardiol. 2013;62(9):851-858.PubMedGoogle ScholarCrossref
26.
Truettner  JS, Alonso  OF, Bramlett  HM, Dietrich  WD.  Therapeutic hypothermia alters microRNA responses to traumatic brain injury in rats.  J Cereb Blood Flow Metab. 2011;31(9):1897-1907.PubMedGoogle ScholarCrossref
27.
Pilotte  J, Dupont-Versteegden  EE, Vanderklish  PW.  Widespread regulation of miRNA biogenesis at the Dicer step by the cold-inducible RNA-binding protein, RBM3.  PLoS One. 2011;6(12):e28446.PubMedGoogle ScholarCrossref
28.
Dresios  J, Aschrafi  A, Owens  GC, Vanderklish  PW, Edelman  GM, Mauro  VP.  Cold stress–induced protein Rbm3 binds 60S ribosomal subunits, alters microRNA levels, and enhances global protein synthesis.  Proc Natl Acad Sci U S A. 2005;102(6):1865-1870.PubMedGoogle ScholarCrossref
29.
Rink  C, Khanna  S.  MicroRNA in ischemic stroke etiology and pathology.  Physiol Genomics. 2011;43(10):521-528.PubMedGoogle ScholarCrossref
30.
Gan  CS, Wang  CW, Tan  KS.  Circulatory microRNA-145 expression is increased in cerebral ischemia.  Genet Mol Res. 2012;11(1):147-152.PubMedGoogle ScholarCrossref
31.
Jeyaseelan  K, Lim  KY, Armugam  A.  MicroRNA expression in the blood and brain of rats subjected to transient focal ischemia by middle cerebral artery occlusion.  Stroke. 2008;39(3):959-966.PubMedGoogle ScholarCrossref
32.
Lakhal  S, Wood  MJA.  Exosome nanotechnology: an emerging paradigm shift in drug delivery: exploitation of exosome nanovesicles for systemic in vivo delivery of RNAi heralds new horizons for drug delivery across biological barriers.  Bioessays. 2011;33(10):737-741.PubMedGoogle ScholarCrossref
33.
Sheinerman  KS, Tsivinsky  VG, Crawford  F, Mullan  MJ, Abdullah  L, Umansky  SR.  Plasma microRNA biomarkers for detection of mild cognitive impairment.  Aging (Albany NY). 2012;4(9):590-605.PubMedGoogle ScholarCrossref
34.
Duttagupta  R, Jiang  R, Gollub  J, Getts  RC, Jones  KW.  Impact of cellular miRNAs on circulating miRNA biomarker signatures.  PLoS One. 2011;6(6):e20769.PubMedGoogle ScholarCrossref
35.
Fichtlscherer  S, De Rosa  S, Fox  H,  et al.  Circulating microRNAs in patients with coronary artery disease.  Circ Res. 2010;107(5):677-684.PubMedGoogle ScholarCrossref
36.
Widera  C, Gupta  SK, Lorenzen  JM,  et al.  Diagnostic and prognostic impact of six circulating microRNAs in acute coronary syndrome.  J Mol Cell Cardiol. 2011;51(5):872-875.PubMedGoogle ScholarCrossref
37.
Karlsson  V, Dankiewicz  J, Nielsen  N,  et al.  Association of gender to outcome after out-of-hospital cardiac arrest: a report from the International Cardiac Arrest Registry.  Crit Care. 2015;19(1):182.PubMedGoogle ScholarCrossref
38.
Tiainen  M, Roine  RO, Pettilä  V, Takkunen  O.  Serum neuron-specific enolase and S-100B protein in cardiac arrest patients treated with hypothermia.  Stroke. 2003;34(12):2881-2886.PubMedGoogle ScholarCrossref
39.
Rundgren  M, Karlsson  T, Nielsen  N, Cronberg  T, Johnsson  P, Friberg  H.  Neuron specific enolase and S-100B as predictors of outcome after cardiac arrest and induced hypothermia.  Resuscitation. 2009;80(7):784-789.PubMedGoogle ScholarCrossref
40.
Oddo  M, Rossetti  AO.  Early multimodal outcome prediction after cardiac arrest in patients treated with hypothermia.  Crit Care Med. 2014;42(6):1340-1347.PubMedGoogle ScholarCrossref
Original Investigation
June 2016

Association of Circulating MicroRNA-124-3p Levels With Outcomes After Out-of-Hospital Cardiac Arrest: A Substudy of a Randomized Clinical Trial

Author Affiliations
  • 1Cardiovascular Research Unit, Luxembourg Institute of Health (LIH), Luxembourg
  • 2Department of Cardiology, Clinical Sciences, Lund University and Skåne University Hospital, Lund, Sweden
  • 3Department of Anaesthesia and Intensive Care Medicine, Centre Hospitalier de Luxembourg, Luxembourg
  • 4Competence Centre for Methodology and Statistics, LIH, Luxembourg
  • 5Department of Cardiology B, The Heart Centre, Rigshospitalet University Hospital, Copenhagen, Denmark
  • 6Department of Intensive Care, University Hospital of Wales, Cardiff
  • 7Department of Intensive Care, Leeuwarden Medical Centrum, Leeuwarden, the Netherlands
  • 8Department of Anesthesia and Intensive Care, Clinical Sciences, Lund University and Skåne University Hospital, Lund, Sweden
  • 9Division of Neurology, Department of Clinical Sciences, Lund University, Lund, Sweden
  • 10Department of Anesthesia and Intensive Care, Clinical Sciences, Lund University and Helsingborg Hospital, Helsingborg, Sweden
 

Copyright 2016 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.

JAMA Cardiol. 2016;1(3):305-313. doi:10.1001/jamacardio.2016.0480
Abstract

Importance  The value of microRNAs (miRNAs) as biomarkers has been investigated in various clinical contexts. Initial small-scale studies suggested that miRNAs might be useful indicators of outcome after cardiac arrest.

Objective  To address the prognostic value of circulating miRNAs in a large cohort of comatose patients with out-of-hospital cardiac arrest.

Design, Setting, and Participants  This substudy of the Target Temperature Management After Cardiac Arrest (TTM) trial, a multicenter randomized, parallel-group, assessor-blinded clinical trial, compared the 6-month neurologic outcomes and survival of patients with cardiac arrest after targeted temperature management at 33°C or 36°C. Five hundred seventy-nine patients who survived the first 24 hours after the return of spontaneous circulation and who had blood samples available for miRNA assessment were enrolled from 29 intensive care units in 9 countries from November 11, 2010, to January 10, 2013. Final follow-up was completed on July 3, 2013, and data were assessed from February 1, 2014, to February 1, 2016.

Interventions  Blood sampling at 48 hours after the return of spontaneous circulation.

Main Outcomes and Measures  The primary end point was poor neurologic outcome at 6 months (cerebral performance category score, 3 [severe neurologic sequelae], 4 [coma], or 5 [death]). The secondary end point was survival until the end of the trial. Circulating levels of miRNAs were measured by sequencing and polymerase chain reaction.

Results  Of the 579 patients (265 men [80.3%]; mean [SD] age, 63 [12] years), 304 patients (52.5%) had a poor neurologic outcome at 6 months. In the discovery phase with short RNA sequencing in 50 patients, the brain-enriched miR-124-3p level was identified as a candidate prognostic variable for neurologic outcomes. In the validation cohort of 529 patients, mean (SD) levels of miR-124-3p were higher in patients with a poor outcome (8408 [12 465] copies/µL) compared with patients with a good outcome (1842 [3025] copies/μL; P < .001). The miR-124-3p level was significantly associated with neurologic outcomes in the univariable analysis (odds ratio, 6.72; 95% CI, 4.53-9.97). In multivariable analyses using logistic regression, miR-124-3p levels were independently associated with neurologic outcomes (odds ratio, 1.62; 95% CI, 1.13-2.32). In Cox proportional hazards models, higher levels of miR-124-3p were significantly associated with lower survival (hazard ratio, 1.63; 95% CI, 1.37-1.93).

Conclusions and Relevance  Levels of miR-124-3p can be used as prognostication tools for neurologic outcome and survival after out-of-hospital cardiac arrest. Thus, miRNA levels may aid in tailoring health care for patients with cardiac arrest.

Trial Registration  clinicaltrials.gov Identifier: NCT01020916

Introduction

Cardiac arrest is a grave condition with a high risk for a poor outcome. Although early in-hospital mortality of patients admitted to the intensive care unit after an out-of-hospital cardiac arrest is mostly owing to hemodynamic failure, more than half of the initial survivors have irreversible neurologic sequelae or die within a few weeks or months.1 Withdrawal of life-sustaining therapies in patients with a presumed poor neurologic prognosis commonly precedes death.2,3 Decisions to withdraw life-sustaining therapies must rely on accurate prognostication tools. However, commonly used methods, such as pathologic patterns on electroencephalograms or neuroimaging, still lack standardization, whereas some clinical signs, such as absent pupillary reflexes to light or absent somatosensory evoked potentials, lack sensitivity.4,5 The biomarker neuron-specific enolase (NSE) is currently recommended in multimodal approaches,6,7 and high serial levels are needed for reliable prediction.8 To increase the accuracy of prognostication and to guide clinical decisions, robust tools are needed. Novel biomarkers would be valuable in consideration of the possibility of adapting health care depending on outcome, maximizing resources in patients likely to survive, and minimizing costly efforts in patients with irreversible neurologic damage.9

MicroRNAs (miRNAs) are small noncoding RNA molecules that regulate gene expression and that are present and stable in the bloodstream.10 In the cardiovascular field, miRNAs have been identified as potential novel tools for personalized health care of patients with heart disease such as myocardial infarction.11 In previous small-scale proof-of-concept studies,12,13 the potential of circulating miRNAs to predict neurologic outcome and death after cardiac arrest was revealed. A large-scale investigation was conducted to confirm these initial findings.

The Target Temperature Management After Cardiac Arrest (TTM) trial14 addressed the potential benefit of targeted temperature management at 33°C vs 36°C for patient outcome after out-of-hospital cardiac arrest.15 In the present ad hoc substudy of the multicenter TTM trial, we addressed the prognostic value of circulating miRNAs for neurologic outcome and mortality 6 months after cardiac arrest.

Box Section Ref ID

Key Points

  • Question Are circulating levels of microRNA (mi-RNA) associated with outcomes after out-of-hospital cardiac arrest?

  • Findings In this substudy of the Target Temperature Management After Cardiac Arrest trial, circulating levels of miR-124-3p measured 48 hours after cardiac arrest were elevated in patients with a poor outcome at 6 months. In multivariable analyses that included neuron-specific enolase, miR-124-3p was independently associated with neurologic outcomes. Higher levels of miR-124-3p were associated with a 2.5-fold increased risk for death.

  • Meaning Levels of miR-124-3p have prognostic value for outcomes after cardiac arrest and may aid in tailoring health care for patients with cardiac arrest.

Methods
Patients

From November 11, 2010, to January 10, 2013, participating centers of the TTM trial enrolled 939 unconscious adult patients admitted to an intensive care unit after an out-of-hospital cardiac arrest of presumed cardiac cause. The primary goal of the trial was to investigate the survival benefit provided by targeted temperature management at 33°C vs 36°C. Details on the TTM trial design, protocol, statistical analysis, and results have been published.15-17 The TTM trial was approved by ethical committees of the 9 participating countries (Czech Republic, Denmark, Italy, Luxembourg, the Netherlands, Norway, Sweden, Switzerland, and United Kingdom). Informed consent was waived or obtained from each participant or their relatives, according to the legislation in each country and in line with the Declaration of Helsinki.18

Of the 36 recruiting centers of the TTM trial, 29 participated in the biobank project and enrolled 700 patients. In the present substudy, we included 579 patients who survived the first 24 hours after the return of spontaneous circulation (ROSC) and who had blood samples available for miRNA assessment (eFigure 1 in the Supplement). This substudy was approved by the steering committee of the TTM trial.

Outcome

The primary end point of this study was poor neurologic outcome as defined by a cerebral performance category (CPC) score of 3 to 5 at 6 months.19 Cerebral performance category scores of 3 or 4 indicate severe neurologic sequelae or coma, respectively, and a CPC score of 5 indicates death. A CPC score of 1 or 2 indicates none or moderate neurologic sequelae, respectively, and is considered a good outcome. Blinded assessment of the CPC score was performed according to the TTM trial protocol.16 The secondary end point was survival until the end of the trial.

miRNA Assessment

Samples used in this study were processed, analyzed, and stored at the Integrated BioBank of Luxembourg in compliance with International Organization for Standards (protocols 9001:2008 and 17025:2005 and NF S96-900:2011) and with the International Society for Biological and Environmental Repositories Best Practices. Circulating levels of miRNAs were first measured using short RNA sequencing in blood samples collected 48 hours after ROSC (discovery phase) (n = 50). Quantitative polymerase chain reaction (PCR) was used subsequently in a validation phase on 48-hour samples (n = 529) and in a time-course experiment with blood samples collected 24, 48, and 72 hours after ROSC (n = 43).

Short RNA Sequencing

Short RNA sequencing of plasma samples was performed at Exiqon Services. Library preparation and sequencing were performed following in-house protocols. Briefly, total RNA was extracted from 400 μL of plasma using a proprietary protocol based on organic extraction, followed by column-based filter clean-up and including a digestion step with proteinase K. Results of the quality check performed for adequate sample preparation are provided in eMethods in the Supplement. Six microliters of total RNA was used to prepare small RNA next-generation sequencing libraries with a multiplex-compatible small RNA library preparation set (NEBNext for Illumina; New England Biolabs, Inc) with some minor modifications. Samples underwent single-end sequencing of 50-nucleotide fragments on a commercially available platform (NextSeq 500; Illumina). Sequencing reads were mapped to miRBase software (version 20; June 2013), which consists of 1871 human precursor sequences and 2772 human mature miRNAs.20 Expression levels of miRNAs are measured as tags per million, which is the number of reads for a particular miRNA divided by the total number of mapped reads and multiplied by 1 million. Differential expression analysis was performed using the EdgeR statistical software package.21 Data normalization was achieved with the trimmed mean of M-values method based on log-fold and absolute gene-wise changes in expression levels between samples.22

Quantitative PCR

Circulating levels of miRNAs were assessed in serum samples as previously described,23 with the following modifications. Briefly, total RNA was extracted from 200 µL of serum samples using a serum/plasma kit and an automated extraction device (miRNeasy and QIAcube, respectively; Qiagen). The synthetic miRNA Caenorhabditis elegans miR-39 (Cel-miR-39; Qiagen) was spiked in samples to correct for extraction efficiency. The RNA underwent reverse transcription (miScript II kit; Qiagen). Complementary DNA was diluted 10-fold before processing in quantitative PCR with the miScript PCR system, including miRNA-specific primer sets (Qiagen). An interrun calibrator was used in each PCR to correct for technical variation between runs. Levels of miRNAs were normalized using the spiked-in control Cel-miR-39 and are expressed as the number of miRNA copies per microliter of serum. Additional details can be found in eMethods and eFigure 2 in the Supplement.

Measurement of NSE Levels

Levels of NSE were measured in serum samples obtained 48 hours after ROSC. Technical details of the NSE level measurement have been previously reported.8 Measurements of miR-124-3p and NSE levels were performed in a core laboratory 6 months after completion of the trial and were thus not available to investigators in the course of the trial.

Statistical Analysis
Demographic and Clinical Data

Follow-up was completed on July 3, 2013, and data were assessed from February 1, 2014, to February 1, 2016. Comparisons of continuous clinical characteristics between 2 groups of patients were performed using the Mann-Whitney test. The χ2 test or the Fisher exact test was used to compare categorical clinical characteristics. P < .05 was considered statistically significant.

Levels of miR-124-3p

Comparisons of miR-124-3p levels between 2 groups of patients were performed using the Mann-Whitney test. The association between miR-124-3p levels and comorbidities was evaluated using logistic regression. Effect of sampling time on miR-124-3p levels was evaluated using 1-way repeated-measures analysis of variance on ranks.

Prediction Analyses

We conducted univariable and multivariable analyses. Logistic regression was used to assess the potential association of miR-124-3p levels after log transformation with neurologic outcome as determined by the dichotomized CPC score at 6 months, with a CPC score of 1 to 2 considered a good neurologic outcome (0 value) and a CPC score of 3 to 5 considered a poor neurologic outcome (1 value). Odds ratios (ORs) were computed for an increase of 1 unit for continuous variables. Continuous variables were centered and scaled to unit to have them on the same scale. The area under the receiver operating characteristic curve (AUC) was used to estimate prediction ability of multivariable models. The procedure was corrected for overfitting using 150-fold bootstrap internal validation,24 which was combined with 10-fold multiple imputation to account for missing data. P values of variables included in the model correspond to the last iteration of the multiple imputation of missing values. The incremental prognostic value of miR-124-3p levels to a clinical model that included demographic and arrest-related factors and NSE levels was evaluated by computation of the integrated discrimination improvement. No marked nonlinear effects, as evaluated using restricted cubic splines, were detected.

Survival Analysis

We used Kaplan-Meier survival analysis and Cox regression to evaluate the association of miR-124-3p levels and survival after log-transformation. In Kaplan-Meier analysis, expression levels of miR-124-3p were divided into quartiles. In Cox regression, the association of miR-124-3p and survival was assessed after adjustment with clinical variables using 10-fold multiple imputation.

Software

Between-group comparisons and survival analysis were performed using the SigmaPlot software (version 12.3; Systat). Prediction analysis was performed using the R statistical environment (version 2.15.2; https://www.r-project.org/) with the ROCR and rms packages.

Results
Patients

A flowchart describing the study design is presented in eFigure 1 in the Supplement. A total of 939 patients were included in the primary analysis of the TTM trial. In the present study, 579 patients who survived the first 24 hours after ROSC were included (265 men [80.3%]; mean [SD] age, 63 [12] years). Blood samples were collected 48 hours after ROSC. This point was chosen for consistency with previous studies12,25 and because NSE levels were maximal 48 hours after cardiac arrest and provided an optimal prognostic value.8 The miRNA study cohort had similar demographic and clinical features to those of the whole TTM cohort (eTable 1 in the Supplement). Three hundred four patients (52.5%) had a good neurologic outcome at 6 months (CPC scores 1-2) and 275 patients (47.5%) had a poor neurologic outcome or died (CPC scores 3-5) (Table 1). Patients with a poor neurologic outcome or who died within the 6-month follow-up period were older and had more comorbidities compared with patients with a good neurologic outcome. The incidence of an initial nonshockable rhythm, time from cardiac arrest to ROSC, initial serum lactate levels, and the frequency of circulatory shock on admission were higher in patients with poor outcomes (Table 1).

Discovery Phase

Short RNA sequencing was used to profile the expression of miRNAs in 2 groups of 25 patients (eFigure 1 in the Supplement). Twenty-five patients had a good neurologic outcome and 25 patients had a poor neurologic outcome or died within the 6 months after cardiac arrest. Consistent with the entire cohort, patients with a poor outcome were older and had a longer time from cardiac arrest to ROSC compared with patients with good outcomes (eTable 2 in the Supplement). However, comorbidities, initial rhythm, lactate levels, and shock on admission were not statistically significantly different between the 2 groups, presumably owing to the small sample size.

After a sample quality check attesting for the suitability of plasma samples for small RNA sequencing (eResults in the Supplement), the sequencing generated a mean of 18.5 million reads per sample (total, 926 million reads for the entire experiment). The data set has been deposited at the Gene Expression Omnibus under the accession number GSE74198. Of the sequencing reads, 84.8% mapped to the human genome and 26.0% mapped to human miRNAs. eFigure 3 in the Supplement depicts the relative contribution of different small RNA species and the percentages of differentially expressed RNAs among different RNA species. Two hundred thirty-six miRNAs were detected in all 50 samples with tags per million greater than or equal to 1. The brain-enriched miR-124-3p was the most differentially expressed miRNA (46-fold change; false discovery rate, 2 × 10−11; P = 7 × 10−14; eFigure 4 in the Supplement) and was selected for validation. Of note, the mean expression level of miR-124-3p was 14 tags per million.

Serum levels of miR-124-3p were measured in 43 of the 50 patients of the discovery phase for whom blood samples at 24, 48, and 72 hours after cardiac arrest were available (eFigure 5 in the Supplement). Although miR-124-3p levels tended to increase over time in the group with poor outcomes, no statistically significant effect of time of sampling was detected (P = .83 in the group with good outcomes and P = .40 in the group with poor outcomes).

Validation Phase

Circulating levels of miR-124-3p were measured using quantitative PCR in a validation cohort of 529 patients from the miRNA cohort (n = 579) after removal of the 50 patients of the discovery phase (eFigure 1 in the Supplement). In this validation cohort, 231 patients (43.7%) were allocated to 33°C treatment and 298 (56.3%) were allocated to 36°C treatment. One hundred fifty-seven patients (52.7%) allocated to 36°C treatment and 121 (52.4%) (52%) allocated to 33°C treatment had a good outcome (P = .98).

Serum levels of miR-124-3p assessed in samples obtained 48 hours after ROSC did not differ between temperature groups (Figure 1A). Patients with a poor outcome had higher levels of miR-124-3p than patients with a good outcome, independently of the targeted temperature regimen (Figure 1B-D). Levels of miR-124-3p were associated with the 6-month neurologic outcome, as determined by dichotomized 6-month CPC scores (1-2 vs 3-5), with an AUC of 0.77 (Figure 2A). This predictive value was comparable in both temperature groups (Figure 2B and C). Patients with high levels of miR-124-3p were at high risk for a poor neurologic outcome (univariate OR, 6.72; 95% CI, 4.53-9.97). Odds ratios were comparable in the 33°C (OR, 6.16; 95% CI, 3.43-11.06) and 36°C (OR, 7.21; 95% CI, 4.23-12.28) temperature groups.

In multivariable analyses that included age, sex, bystander cardiopulmonary resuscitation, first monitored rhythm, time from cardiac arrest to ROSC, initial serum lactate levels, circulatory shock on admission, 48-hour NSE levels, and temperature regimen, levels of miR-124-3p were independent predictors of neurologic outcome (Table 2). Age, NSE levels, and first monitored rhythm also had a significant prognostic value. The multivariable model was associated with neurologic outcome with an AUC of 0.90. Temperature regimen was not associated with neurologic outcomes, in accordance with previous data from the main TTM trial.15

To determine the added value of miR-124-3p levels, we compared the AUCs of a model fitted with clinical variables and NSE levels only to the preceding model that included clinical variables, NSE levels, and miR-124-3p levels. Inclusion of miR-124-3p levels slightly improved prognostic accuracy as attested by a modest, albeit significant, increase of the AUC from 0.89 to 0.90 (P = .02). However, miR-124-3p levels did not reclassify a significant number of patients as attested by an integrated discrimination improvement of 0.005 (95% CI, −0.003 to 0.013). When NSE levels were omitted from the prediction models, miR-124-3p levels provided a most prominent improvement of prognostic ability, with an AUC of 0.78 and 0.84 for the models without and with miR-124-3p levels, respectively (P < .001). In this case, the integrated discrimination improvement generated by miR-124-3p levels was 0.11 (95% CI, 0.08-0.14).

Survival Analysis

The association of circulating levels of miR-124-3p measured 48 hours after ROSC with survival until the end of the trial was evaluated using Cox proportional hazards models. Maximum follow-up time was 956 days after cardiac arrest. Median time from cardiac arrest to death was 275 days. Combined with clinical and demographic variables, miR-124-3p levels had significant prognostic value for a shorter survival (Table 2), as did age, first monitored rhythm, initial serum lactate levels, and NSE levels. Targeted temperature regimen was not associated with survival, consistent with a previous report.15 Inclusion of miR-124-3p levels improved the predictive value with an integrated discrimination improvement of 0.06 (95% CI, 0.02-0.10). In Kaplan-Meier analysis, miR-124-3p levels were strongly associated with shorter survival (Figure 2D), even after adjustment for age. Overall, patients with high levels of miR-124-3p 48 hours after ROSC were at high risk for death.

Discussion

In this predefined substudy of the multicenter TTM trial, we addressed the prognostic value of circulating miRNAs after cardiac arrest. We report that the miR-124-3p level is associated with neurologic outcome and death, independently of the targeted temperature management regimen.

Animal experiments26 and in vitro studies27,28 have shown that mild induced hypothermia affects the expression of some miRNAs in the brain. In our study, circulating levels of miR-124-3p were not affected by targeted temperature. This finding is consistent with the observed absence of effect of hypothermia on the predictive value of miR-124-3p levels and with the equal effect of targeted temperature management at 33°C or 36°C to protect the brain from severe neurologic sequelae induced by hypoxia and ischemia.15

Patients with high levels of miR-124-3p were at high risk for poor neurologic outcome and death. This outcome confirms the results of a previous study13 and is consistent with the concept that circulating levels of brain-enriched miRNAs reflect neurologic damage. Several other studies support this concept. First, the blood-brain barrier is disrupted after cerebral ischemia,29 allowing the release of brain-enriched miRNAs into the bloodstream.30,31 Second, exosomes carrying miRNAs have the ability to cross the blood-brain barrier.32 Third, brain-enriched miRNAs have been identified in the blood samples of patients at an early stage of neurodegenerative disease.33 Furthermore, in the present study, levels of miR-124-3p were positively correlated with the time from cardiac arrest to ROSC and with NSE levels, but not with age. Because most patients with cardiac arrest die of irreversible brain injury, the prognostic ability of brain-enriched miRNAs for death after cardiac arrest is biologically plausible. We expect that patients with high serum levels of brain-enriched miR-124-3p, attesting to an extensive brain damage, are at high risk for death.

Sex differences in circulating levels of miRNAs have been reported in some34-36 but not all11 miRNA studies. We did not detect different levels of miR-124-3p between male and female patients. Of note, sex was not associated with outcome in the present substudy, in the main TTM trial,15 and in the previous proof-of-concept study.12 In a recent report from the International Cardiac Arrest Registry,37 men were more likely than women to survive after out-of-hospital cardiac arrest despite similar neurologic outcomes. Thus, whether sex is an important determinant of outcome after cardiac arrest is still a matter of debate.

Identification of novel biomarkers associated with outcomes after cardiac arrest is important because currently available biomarkers for prognostication are insensitive when aiming for a low false-positive rate and may be hindered by confounders such as hemolysis. Previous investigations8,38,39 demonstrated the potential of brain-derived proteins such as NSE and S-100β to identify patients at high risk for a poor outcome. Combined determination of clinical examination, biomarkers, and neurophysiologic variables has shown some added prognostic value.25,40 Herein, we have shown that circulating miR-124-3p can aid in the prediction of outcome after cardiac arrest. The observation that miR-124-3p levels do not provide an added predictive value to a clinical model that includes NSE is consistent with the fact that NSE and miR-124-3p levels reflect cerebral damage. Other miRNAs indicative of multiple organ injury may have an added value and should be considered in further investigations.

From a clinical perspective, our findings are appealing because diagnostic assays to measure miRNA levels are being implemented.11 These assays require a low volume of blood, are sensitive, can be automated and multiplexed, and may thus be cost-effective. However, the added value in a prognostic algorithm involving clinical examination, neurophysiology, and brain imaging remains to be established. In future investigations, the focus should be on the added value of miRNA levels for early prognostication. Because high, serial levels of NSE are necessary for reliable prediction,8 one may expect that miRNAs, which are small molecules that easily cross the blood-brain barrier, may be detectable in the blood earlier than NSE. In addition, brain-enriched miR-124-3p might also serve as a biomarker in other rapidly progressive neurodegenerative or neurovascular disorders such as stroke or subarachnoid hemorrhage.

This study has a number of strengths and limitations. Our examination of miR-124-3p level as a prognostic factor was a predefined substudy of the largest prospective multicenter randomized clinical trial—to our knowledge—to address the therapeutic value of 2 target temperature regimens in patients with cardiac arrest. In addition, assessment of miR-124-3p levels was performed in a single laboratory to maximize consistency of sample handling and processing and measurement of miRNA levels. We have implemented a computational method to determine the absolute concentrations of miR-124-3p, whereas most past miRNA studies reported relative concentrations.

As far as limitations are concerned, the predictive value of only 1 miRNA is reported, and other miRNAs may also have a prognostic value. We must acknowledge that short RNA sequencing of plasma samples is prone to multiple biases that may prevent the discovery of novel candidates. Also, the study was performed in a subgroup of patients of the TTM trial for whom blood samples were available. However, demographic and clinical features of this subgroup are comparable to those of the whole cohort. The discovery phase with sequencing was undertaken in plasma samples, whereas the validation phase with PCR was achieved in serum samples. Although this process might appear inconsistent, the ability to confirm initial findings obtained in plasma samples in additional serum samples validates the suitability of both sample types to be used for miRNA testing. No mechanistic link between miR-124-3p levels and outcome are provided in this biomarker study, although we expect that patients with high circulating levels of brain-enriched miR-124-3p, reflecting extensive brain injury, are at high risk for poor outcomes. Finally, serial measurements were performed in a subgroup of 43 patients, and the best time points for outcome prognostication remain to be determined.

Conclusions

The mRNA-124-3p level is associated with neurologic outcome and survival after cardiac arrest. This finding motivates future research to determine the additive value of measurement of miR-124-3p levels for early prognostication after cardiac arrest.

Back to top
Article Information

Corresponding Author: Yvan Devaux, PhD, Cardiovascular Research Unit, Luxembourg Institute of Health, 84 Val Fleuri, L-1526 Luxembourg (yvan.devaux@lih.lu).

Accepted for Publication: January 25, 2016.

Published Online: May 18, 2016. doi:10.1001/jamacardio.2016.0480.

Author Contributions: Dr Devaux had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Devaux, Stammet, Collignon, Gilje, Gidlöf, Hassager, Kuiper, Friberg, Erlinge, Nielsen.

Acquisition, analysis, or interpretation of data: Devaux, Dankiewicz, Salgado-Somoza, Stammet, Collignon, Zhang, Vausort, Hassager, Wise, Kuiper, Cronberg, Erlinge, Nielsen.

Drafting of the manuscript: Devaux, Collignon.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Devaux, Collignon, Gidlöf.

Obtained funding: Devaux, Cronberg, Erlinge, Nielsen.

Administrative, technical, or material support: Devaux, Stammet, Gilje, Vausort, Cronberg, Erlinge.

Study supervision: Devaux, Cronberg, Nielsen.

Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Wise reports serving as a paid member of an advisory board to BARD Medical, Inc. No other disclosures were reported.

Funding/Support: This study was supported by the main TTM trial; grant C14/BM/8225223 from the Ministry of Culture, Higher Education and Research, National Research Fund of Luxembourg; the Swedish Heart Lung Foundation; Arbetsmarknadens försäkringsaktiebolag (AFA) Insurance Foundation; the Swedish Research Council; Region Skåne research support; governmental funding of clinical research within the Swedish National Health Services; Thelma Zoega Foundation; Krapperup Foundation; Thure Carlsson Foundation; Hans-Gabriel and Alice Trolle-Wachtmeister Foundation for Medical Research; Skåne University Hospital; TrygFonden; and the European Clinical Research Infrastructures Network.

Role of the Funder/Sponsor: The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Group Members: The following Target Temperature Management After Cardiac Arrest (TTM) trial investigators and centers participated in this study: Niklas Nielsen, Department of Anesthesiology and Intensive Care, Helsingborg Hospital, Helsingborg, Sweden; Dr Nielsen, Tobias Cronberg, David Erlinge, G. Lilja, M. Rundgren, and Hans Friberg, Department of Clinical Sciences, and Dr Erlinge, Department of Cardiology, Lund University, Lund, Sweden; Dr Cronberg, Department of Neurology, M. Rundgren and Dr Friberg, Department of Anesthesiology and Intensive Care, and G. Lilja, Department of Rehabilitation Medicine, Skåne University Hospital, Lund; C. Rylander, Department of Anesthesiology and Intensive Care, Sahlgrenska University Hospital, Gothenburg, Sweden; Jørn Wetterslev and P. Winkel, Copenhagen Trial Unit, Center of Clinical Intervention Research, Christian Hassager, Jesper Kjaergaard, S. Boesgaard, J. Bro-Jeppesen, L. Køber, and J. E. Møller, Departments of Cardiology and Cardiothoracic Anesthesiology, and Michael Wanscher, the Heart Center, Copenhagen University Hospital Rigshospitalet, Copenhagen, Denmark; Yvan Gasche, the Department of Anesthesiology, Pharmacology, and Intensive Care, Geneva University Hospital, Geneva, Switzerland); Janneke Horn, N. P. Juffermans, and Michael Kuiper, Department of Intensive Care, Academic Medical Center, Amsterdam, the Netherlands; Jan Hovdenes and J. F. Bugge, Department of Anesthesiology, Oslo University Hospital, Rikshospitalet, Oslo, Norway; Dr Kuiper and M. Koopmans, Department of Intensive Care, Medical Center Leeuwarden, Leeuwarden, the Netherlands; Thomas Pellis, Intensive Care Unit, Santa Maria degli Angeli, Pordenone, Italy; I. Brunetti, Department of Intensive Care, Istituto di Ricovero e Cura a Carattere Scientifico San Martino, Istituto Scientifico Tumori, University of Genoa, Genoa, Italy; Pascal Stammet, MD, and C. Werer, Department of Anesthesiology and Intensive Care, Centre Hospitalier de Luxembourg, Luxembourg; Matthew P. Wise, and C. D. Hingston, Adult Critical Care, University Hospital of Wales, Cardiff, Wales; Anders Åneman, Department of Intensive Care, Liverpool Hospital, Sydney, Australia; N. Al-Subaie, Department of Intensive Care, St George’s Hospital, London, England; J. Langørgen, Department of Heart Diseases, Haukeland University Hospital, Bergen, Norway; and O. Smid, Second Department of Internal Medicine, Cardiology and Angiology, General University Hospital in Prague, Prague, Czech Republic. Members of the steering committee include Drs Nielsen (chairperson and chief investigator), Cronberg, Erlinge, Friberg (senior investigator), Hassager, Horn, Hovdenes, Kjaergaard, Kuiper, Gasche, Pellis, Stammet, Wanscher, Wetterslev, Wise, and Åneman.

Additional Information: The following authors are members of the Cardiolinc Network (http://www.cardiolinc.org/): Yvan Devaux, PhD, Antonio Salgado-Somoza, PhD, Pascal Stammet, MD, Lu Zhang, MSc, Melanie Vausort, MSc, Hans Friberg, MD, PhD, and David Erlinge, PhD.

Additional Contributions: François Massart, BSc, Cardiovascular Research Unit, Luxembourg Institute of Health, Luxembourg, provided technical help, for which he received no compensation. We thank all TTM trial investigators involved in the biomarker studies.

References
1.
Lemiale  V, Dumas  F, Mongardon  N,  et al.  Intensive care unit mortality after cardiac arrest: the relative contribution of shock and brain injury in a large cohort.  Intensive Care Med. 2013;39(11):1972-1980.PubMedGoogle ScholarCrossref
2.
Dragancea  I, Rundgren  M, Englund  E, Friberg  H, Cronberg  T.  The influence of induced hypothermia and delayed prognostication on the mode of death after cardiac arrest.  Resuscitation. 2013;84(3):337-342.PubMedGoogle ScholarCrossref
3.
Laver  S, Farrow  C, Turner  D, Nolan  J.  Mode of death after admission to an intensive care unit following cardiac arrest.  Intensive Care Med. 2004;30(11):2126-2128.PubMedGoogle ScholarCrossref
4.
Sandroni  C, Cavallaro  F, Callaway  CW,  et al.  Predictors of poor neurological outcome in adult comatose survivors of cardiac arrest: a systematic review and meta-analysis, part 2: patients treated with therapeutic hypothermia.  Resuscitation. 2013;84(10):1324-1338.PubMedGoogle ScholarCrossref
5.
Friberg  H, Cronberg  T, Dünser  MW, Duranteau  J, Horn  J, Oddo  M.  Survey on current practices for neurological prognostication after cardiac arrest.  Resuscitation. 2015;90:158-162.PubMedGoogle ScholarCrossref
6.
Sandroni  C, Cariou  A, Cavallaro  F,  et al.  Prognostication in comatose survivors of cardiac arrest: an advisory statement from the European Resuscitation Council and the European Society of Intensive Care Medicine.  Resuscitation. 2014;85(12):1779-1789.PubMedGoogle ScholarCrossref
7.
Nolan  JP, Soar  J, Cariou  A,  et al.  European Resuscitation Council and European Society of Intensive Care Medicine Guidelines for Post-resuscitation Care 2015: section 5 of the European Resuscitation Council Resuscitation Guidelines 2015.  Resuscitation. 2015.PubMedGoogle Scholar
8.
Stammet  P, Collignon  O, Hassager  C,  et al; TTM-Trial Investigators.  Neuron-specific enolase as a predictor of death or poor neurological outcome after out-of-hospital cardiac arrest and targeted temperature management at 33°C and 36°C.  J Am Coll Cardiol. 2015;65(19):2104-2114.PubMedGoogle ScholarCrossref
9.
Devaux  Y, Stammet  P, Friberg  H,  et al; Biomarker Subcommittee of TTM trial (Target Temperature Management After Cardiac Arrest, NCT01020916).  MicroRNAs: new biomarkers and therapeutic targets after cardiac arrest?  Crit Care. 2015;19(1):54.PubMedGoogle ScholarCrossref
10.
Mitchell  PS, Parkin  RK, Kroh  EM,  et al.  Circulating microRNAs as stable blood-based markers for cancer detection.  Proc Natl Acad Sci U S A. 2008;105(30):10513-10518.PubMedGoogle ScholarCrossref
11.
Goretti  E, Wagner  DR, Devaux  Y.  miRNAs as biomarkers of myocardial infarction: a step forward towards personalized medicine?  Trends Mol Med. 2014;20(12):716-725.PubMedGoogle ScholarCrossref
12.
Stammet  P, Goretti  E, Vausort  M, Zhang  L, Wagner  DR, Devaux  Y.  Circulating microRNAs after cardiac arrest.  Crit Care Med. 2012;40(12):3209-3214.PubMedGoogle ScholarCrossref
13.
Gilje  P, Gidlöf  O, Rundgren  M,  et al.  The brain-enriched microRNA miR-124 in plasma predicts neurological outcome after cardiac arrest.  Crit Care. 2014;18(2):R40.PubMedGoogle ScholarCrossref
14.
ClinicalTrials.gov. Target Temperature Management After Cardiac Arrest. NCT01020916. https://clinicaltrials.gov/ct2/show/NCT01020916. Accessed April 5, 2016.
15.
Nielsen  N, Wetterslev  J, Cronberg  T,  et al; TTM Trial Investigators.  Targeted temperature management at 33°C vs 36°C after cardiac arrest.  N Engl J Med. 2013;369(23):2197-2206.PubMedGoogle ScholarCrossref
16.
Nielsen  N, Wetterslev  J, al-Subaie  N,  et al.  Target temperature management after out-of-hospital cardiac arrest: a randomized, parallel-group, assessor-blinded clinical trial—rationale and design.  Am Heart J. 2012;163(4):541-548.PubMedGoogle ScholarCrossref
17.
Nielsen  N, Winkel  P, Cronberg  T,  et al.  Detailed statistical analysis plan for the target temperature management after out-of-hospital cardiac arrest trial.  Trials. 2013;14(1):300.PubMedGoogle ScholarCrossref
18.
World Medical Association.  World Medical Association Declaration of Helsinki: ethical principles for medical research involving human subjects.  JAMA. 2013;310(20):2191-2194.PubMedGoogle ScholarCrossref
19.
Booth  CM, Boone  RH, Tomlinson  G, Detsky  AS.  Is this patient dead, vegetative, or severely neurologically impaired? assessing outcome for comatose survivors of cardiac arrest.  JAMA. 2004;291(7):870-879.PubMedGoogle ScholarCrossref
20.
Kozomara  A, Griffiths-Jones  S.  miRBase: integrating microRNA annotation and deep-sequencing data.  Nucleic Acids Res. 2011;39(Database issue):D152-D157.PubMedGoogle ScholarCrossref
21.
Robinson  MD, McCarthy  DJ, Smyth  GK.  edgeR: a Bioconductor package for differential expression analysis of digital gene expression data.  Bioinformatics. 2010;26(1):139-140.PubMedGoogle ScholarCrossref
22.
Robinson  MD, Oshlack  A.  A scaling normalization method for differential expression analysis of RNA-seq data.  Genome Biol. 2010;11(3):R25.PubMedGoogle ScholarCrossref
23.
Devaux  Y, Vausort  M, Goretti  E,  et al.  Use of circulating microRNAs to diagnose acute myocardial infarction.  Clin Chem. 2012;58(3):559-567.PubMedGoogle ScholarCrossref
24.
Harrell  FE.  Regression Modeling Strategies. New York, NY: Springer Science & Business Media; 2001.
25.
Stammet  P, Wagner  DR, Gilson  G, Devaux  Y.  Modeling serum level of s100β and bispectral index to predict outcome after cardiac arrest.  J Am Coll Cardiol. 2013;62(9):851-858.PubMedGoogle ScholarCrossref
26.
Truettner  JS, Alonso  OF, Bramlett  HM, Dietrich  WD.  Therapeutic hypothermia alters microRNA responses to traumatic brain injury in rats.  J Cereb Blood Flow Metab. 2011;31(9):1897-1907.PubMedGoogle ScholarCrossref
27.
Pilotte  J, Dupont-Versteegden  EE, Vanderklish  PW.  Widespread regulation of miRNA biogenesis at the Dicer step by the cold-inducible RNA-binding protein, RBM3.  PLoS One. 2011;6(12):e28446.PubMedGoogle ScholarCrossref
28.
Dresios  J, Aschrafi  A, Owens  GC, Vanderklish  PW, Edelman  GM, Mauro  VP.  Cold stress–induced protein Rbm3 binds 60S ribosomal subunits, alters microRNA levels, and enhances global protein synthesis.  Proc Natl Acad Sci U S A. 2005;102(6):1865-1870.PubMedGoogle ScholarCrossref
29.
Rink  C, Khanna  S.  MicroRNA in ischemic stroke etiology and pathology.  Physiol Genomics. 2011;43(10):521-528.PubMedGoogle ScholarCrossref
30.
Gan  CS, Wang  CW, Tan  KS.  Circulatory microRNA-145 expression is increased in cerebral ischemia.  Genet Mol Res. 2012;11(1):147-152.PubMedGoogle ScholarCrossref
31.
Jeyaseelan  K, Lim  KY, Armugam  A.  MicroRNA expression in the blood and brain of rats subjected to transient focal ischemia by middle cerebral artery occlusion.  Stroke. 2008;39(3):959-966.PubMedGoogle ScholarCrossref
32.
Lakhal  S, Wood  MJA.  Exosome nanotechnology: an emerging paradigm shift in drug delivery: exploitation of exosome nanovesicles for systemic in vivo delivery of RNAi heralds new horizons for drug delivery across biological barriers.  Bioessays. 2011;33(10):737-741.PubMedGoogle ScholarCrossref
33.
Sheinerman  KS, Tsivinsky  VG, Crawford  F, Mullan  MJ, Abdullah  L, Umansky  SR.  Plasma microRNA biomarkers for detection of mild cognitive impairment.  Aging (Albany NY). 2012;4(9):590-605.PubMedGoogle ScholarCrossref
34.
Duttagupta  R, Jiang  R, Gollub  J, Getts  RC, Jones  KW.  Impact of cellular miRNAs on circulating miRNA biomarker signatures.  PLoS One. 2011;6(6):e20769.PubMedGoogle ScholarCrossref
35.
Fichtlscherer  S, De Rosa  S, Fox  H,  et al.  Circulating microRNAs in patients with coronary artery disease.  Circ Res. 2010;107(5):677-684.PubMedGoogle ScholarCrossref
36.
Widera  C, Gupta  SK, Lorenzen  JM,  et al.  Diagnostic and prognostic impact of six circulating microRNAs in acute coronary syndrome.  J Mol Cell Cardiol. 2011;51(5):872-875.PubMedGoogle ScholarCrossref
37.
Karlsson  V, Dankiewicz  J, Nielsen  N,  et al.  Association of gender to outcome after out-of-hospital cardiac arrest: a report from the International Cardiac Arrest Registry.  Crit Care. 2015;19(1):182.PubMedGoogle ScholarCrossref
38.
Tiainen  M, Roine  RO, Pettilä  V, Takkunen  O.  Serum neuron-specific enolase and S-100B protein in cardiac arrest patients treated with hypothermia.  Stroke. 2003;34(12):2881-2886.PubMedGoogle ScholarCrossref
39.
Rundgren  M, Karlsson  T, Nielsen  N, Cronberg  T, Johnsson  P, Friberg  H.  Neuron specific enolase and S-100B as predictors of outcome after cardiac arrest and induced hypothermia.  Resuscitation. 2009;80(7):784-789.PubMedGoogle ScholarCrossref
40.
Oddo  M, Rossetti  AO.  Early multimodal outcome prediction after cardiac arrest in patients treated with hypothermia.  Crit Care Med. 2014;42(6):1340-1347.PubMedGoogle ScholarCrossref
×