Conflict of Interest Declaration
The ILCOR Continuous Evidence Evaluation process is guided by a rigorous ILCOR Conflict of Interest policy. The following Task Force members and other authors were recused from the discussion as they declared a conflict of interest: none applicable
The following Task Force members and other authors declared an intellectual conflict of interest and this was acknowledged and managed by the Task Force Chairs and Conflict of Interest committees: Tobias Cronberg, is co-author of some of the included studies in the present review. He was excluded from the bias assessment of these studies.
CoSTR Citation
Sandroni C, Cacciola S, Cronberg T, D’Arrigo S, Hoedemaekers CWE, Kamps M, Nolan JP, Böttiger BW, Andersen LW , Callaway CW, Deakin CD, Donnino MW, Drennan I, Hsu C, Morley PM, Nicholson TC, O’Neil BJ, Neumar RW, Paiva EF, Parr MJ, Reynolds JC, Wang TL, Welsford M, Berg KM, Soar J. Imaging for prognostication. Consensus on Science with Treatment Recommendations [Internet] Brussels, Belgium: International Liaison Committee on Resuscitation (ILCOR) Advanced Life Support Task Force, 2020 Jan 1. Available from: http://ilcor.org.
Methodological preamble
The continuous evidence evaluation process for the production of Consensus on Science with Treatment Recommendations (CoSTR) started with a systematic review of prognostication after cardiac arrest (Sandroni C 2020 – PROSPERO: CRD 420 1914 1169) conducted by a systematic review team with involvement of clinical content experts from the ILCOR ALS Task Force.
Systematic review
Sandroni C et al. Imaging for prognostication in comatose survivors of cardiac arrest. In preparation.
PICOST
The PICOST (Population, Intervention, Comparator, Outcome, Study Designs and Timeframe)
Population: Adults who are comatose after resuscitation from cardiac arrest (either in-hospital or out-of-hospital), regardless of target temperature.
Intervention: Imaging studies assessed within one week from cardiac arrest.
Comparator: none.
Outcome: Prediction of poor neurological outcome defined as Cerebral Performance Categories (CPC) 3-5 or modified Rankin Score (mRS) 4-6 at hospital discharge/1 month or later.
Study Design: Prognostic accuracy studies where the 2 x 2 contingency table (i.e., the number of true/false negatives and positives for prediction of poor outcome) was reported, or where those variables could be calculated from reported data, are eligible for inclusion. Unpublished studies, reviews, case reports, case series, studies including less than 10 patients, letters, editorials, conference abstracts, and studies published in abstract form were excluded.
Timeframe: In 2015, an ILCOR evidence review identified four categories of predictors of neurological outcome after cardiac arrest, namely clinical examination, biomarkers, electrophysiology and imaging. In the last four years, several studies have been published and new predictors have been identified, therefore the topic needs an update.
The most recent search of the previous systematic reviews on neuroprognostication was launched on May 31, 2013. We searched studies published from January 1, 2013 onwards.
PROSPERO: CRD 420 1914 1169
Consensus on science
Grey matter to white matter ratio (GWR)
GWR-AVERAGE (GWR-AVG)
GWR-AVG was investigated in seven observational studies [Jeon 2017 118; Kim 2013 57; Kim 2014 1121; Kim 2018 33; Lee 2017 1628; Wang 2018 599; Youn 2017 120].
In four studies [Jeon 2017 118, 39 pts; Kim 2013 57, 51 pts; Kim 2014 1121, 91 pts; Kim 2018 33, 174 pts] GWR-AVG ≤1.23 within 6h from ROSC predicted poor neurological outcome from hospital discharge to 6 months with 100% specificity and sensitivity ranging from 13.3% to 83.8% (certainty of evidence from low to very-low).
In one study [Lee 2017 1628, 67 pts] GWR-AVG ≤1.13 at 124.5±59.9 min from ROSC predicted poor neurological outcome at 1 month with 85% specificity and 29.8% sensitivity (very-low certainty of evidence).
In one study [Youn 2017 120, 240 pts] GWR-AVG ≤1.077 within 24h from ROSC predicted poor neurological outcome at hospital discharge with 100% specificity and 15.6% sensitivity (very-low certainty of evidence).
In one study [Wang 2018 599, 58 pts] GWR-AVG ≤1.14 within 72h from ROSC predicted poor neurological outcome at hospital discharge with 100% specificity and 38.1% sensitivity (very-low certainty of evidence).
GWR-Basal Ganglia (GWR-BG)
GWR-BG was investigated in four observational studies [Kim 2013 57; Scarpino 2018 114; Scarpino 2019 115; Wang 2018 599].
In one study [Kim 2013 57, 51 pts] GWR-BG ≤1.12 within 1h from ROSC predicted poor neurological outcome at hospital discharge with 100% specificity and 3.3% sensitivity (very-low certainty of evidence).
In two studies [Scarpino 2018 114, 183 pts; Scarpino 2019 115, 346 pts] GWR-BG ≤1.21 within 24h from ROSC predicted poor neurological outcome at 6 months with 100% specificity and sensitivity ranging from 41.8% to 42.1% (certainty of evidence from moderate to very low).
In one study [Wang 2018 599, 58 pts] GWR-BG ≤1.12 within 72h from ROSC predicted poor neurological outcome at hospital discharge with 100% specificity and 28.6% sensitivity (very-low certainty of evidence).
GWR Putamen/Corpus Callosum (P/CC)
GWR-P/CC was investigated in three observational studies [Lee 2013 1387; Lee 2018 37; Jeon 2017, 21].
In two studies [Lee 2013 1387, 186 pts; Jeon 2017 21, 39 pts] GWR-P/CC ≤1.17 within 6h from ROSC predicted poor neurological outcome from hospital discharge to 6 months with 100% specificity and sensitivity ranging from 31.3% to 52.9% (very-low certainty of evidence).
In one study [Lee 2018 37, 258 pts] GWR-P/CC ≤0.91 within 24h from ROSC predicted poor neurological outcome at 6 months with 100% specificity and 1.7% sensitivity (very-low certainty of evidence).
GWR-Simplified (GWR-SI: Putamen/Posterior limb of internal capsule).
GWR-SI was investigated in one observational study [Wang 2018 1599].
In one study [Wang 2018 1599, 58 pts] GWR-SI ≤1.1 within 72h from ROSC predicted poor neurological outcome at hospital discharge with 100% sensitivity and 28.6% sensitivity (very-low certainty of evidence).
GWR Caudate Nucleus/Posterior limb of internal capsule (CN/PIC)
GWR-CN/PIC was investigated in two observational studies [Lee 2013, 186 pts; Jeon 2017, 39 pts].
In two studies [Lee 2013 1387, 186 pts; Jeon 2017 21, 39 pts] GWR-CN/PIC ≤1.15 within 6h from ROSC predicted poor neurological outcome from hospital discharge to 6 months with 100% specificity and sensitivity ranging from 19.8% to 40.6% (very-low certainty of evidence).
GWR cerebrum
GWR-cerebrum was investigated in two observational studies [Kim 2013 (a) 57; Wang 2018 1599].
In one study [Kim 2013 57, 51 pts] GWR-cerebrum ≤1.12 within 1h from ROSC predicted poor neurological outcome at hospital discharge with 100% specificity and 20% sensitivity (very-low certainty of evidence).
In one study [Wang 2018 599, 58 pts] GWR-cerebrum ≤1.09 within 72h from ROSC predicted poor neurological outcome at hospital discharge with 100% specificity and 28.6% sensitivity (very-low certainty of evidence).
GWR Thalamus/Corpus Callosum (GWR-T/CC)
GWR-T/CC was investigated in one observational study [Jeon 2017 118, 39 pts].
In this study GWR-T/CC ≤1.13 at median time of 90 (IQR 52–150) min predicted poor neurological outcome at 6 months with 100% specificity and 50% sensitivity (very-low certainty of evidence).
GWR Caudate nucleus /Corpus callosum (GWR-CN/CC)
GWR-CN/CC was investigated in one observational study [Jeon 2017 118, 39 pts].
In this study GWR-CN/CC ≤1.15 at median time of 90 (IQR 52–150) min predicted poor neurological outcome at 6 months with 100% specificity and 46.9% sensitivity (very-low certainty of evidence).
GWR in cardiac vs. non-cardiac etiology
In one study including CA with cardiac aetiology [Lee 2015 46, 283 pts] GWR-AVR ≤1.13 at 50 (IQR 26-107) min from ROSC predicted poor neurological outcome at hospital discharge with 100% specificity and 3.5% sensitivity (very-low certainty of evidence).
In one study including CA with non-cardiac aetiology [Lee 2016 1583, 164 pts] GWR-AVR ≤1.22 at 67 (IQR 29-115) min from ROSC predicted poor neurological outcome at hospital discharge with 100% specificity and 28.3% sensitivity (very-low certainty of evidence).
In one study including CA with cardiac aetiology [Lee 2015 46, 283 pts] GWR-BG ≤1.11 at 50 (IQR 26-107) min from ROSC predicted poor neurological outcome at hospital discharge with 100% specificity and 3.5% sensitivity (very-low certainty of evidence).
In one study including CA with non-cardiac aetiology [Lee 2016 1583, 164 pts] GWR-BG ≤1.17 at 67 (IQR 29-115) min from ROSC predicted poor neurological outcome at hospital discharge with 100% specificity and 26.2% sensitivity (very-low certainty of evidence).
In one study including CA with cardiac aetiology [Lee 2015 46, 283 pts] GWR-P/CC ≤1.107 at 50 (IQR 26-107) min from ROSC predicted poor neurological outcome at hospital discharge with 100% specificity and 5.6% sensitivity (very-low certainty of evidence).
In one study including CA with non-cardiac aetiology [Lee 2016 1583, 164 pts] GWR-P/CC ≤1.2 at 67 (IQR 29-115) min from ROSC predicted poor neurological outcome at hospital discharge with 100% specificity and 43.4% sensitivity (very-low certainty of evidence).
In one study including CA with cardiac aetiology [Lee 2015 46, 283 pts;] GWR-SI ≤1.06 at 50 (IQR 26-107) min from ROSC predicted poor neurological outcome at hospital discharge with 100% sensitivity and 3.5% sensitivity (very-low certainty of evidence).
In one study including CA with non-cardiac aetiology [Lee 2016 1583, 164 pts] GWR-SI ≤1.12 at 67 (IQR 29-115) min from ROSC predicted poor neurological outcome at hospital discharge with 100% sensitivity and 9.7% sensitivity (very-low certainty of evidence).
In one study including CA with cardiac aetiology [Lee 2015 46, 283 pts] GWR-CN/PIC ≤1.094 at 50 (26-107) min from ROSC predicted poor neurological outcome at hospital discharge with 100% specificity and 3.5% sensitivity (very-low certainty of evidence).
In one study including CA with non-cardiac aetiology [Lee 2016 1583, 164 pts] GWR-CN/PIC ≤1.138 at 67 (29-115) min from ROSC predicted poor neurological outcome at hospital discharge with 100% specificity and 20% sensitivity (very-low certainty of evidence).
In one study including CA with cardiac aetiology [Lee 2015 46, 283 pts] GWR-cerebrum ≤1.15 at 50 (26-107) min from ROSC predicted poor neurological outcome at hospital discharge with 100% specificity and 4.2% sensitivity (very-low certainty of evidence).
In one study including CA with non-cardiac aetiology [Lee 2016 1583, 164 pts] GWR-cerebrum ≤1.2 at 67 (29-115) min from ROSC predicted poor neurological outcome at hospital discharge with 100% specificity and 11% sensitivity (very-low certainty of evidence).
Diffusion-weighted imaging (DWI)
DWI was investigated in five observational studies [Greer 2013 1546; Jang 2019 142; Jeon 2017 118; Kim 2018 33; Ryoo 2015 2370].
In one study [Jeon 2017 118, 39 pts] high signal intensity on DWI-MRI within 6h from ROSC predicted poor neurological outcome at 6 months with 100% specificity and 81.3% sensitivity (very-low certainty of evidence).
In four studies [Greer 2013 1546, 80 pts; Jang 2019 142, 39 pts, Kim 2018 33, 133 pts; Ryoo 2015 2370, 172 pts] positive findings on DWI-MRI within 5 days predicted poor neurological outcome from hospital discharge to 6 months with specificity ranging from 55.7% to 100% and sensitivity ranging from 26.9% to 92.6% (very-low certainty of evidence).
Apparent diffusion coefficient (ADC)
ADC was investigated in three studies [Moon 2018 36; Kim 2013 1393; Hirsch 2019 in press].
In one study [Moon 2018 36, 44 pts] mean ADC ≤726× 10−6 mm2/s at <48h predicted poor neurological outcome at 6 months with 100% specificity and 44% sensitivity (very-low certainty of evidence).
In one study [Moon 2018 36, 66 pts] mean ADC ≤627× 10−6 mm2/s at 48h-7days predicted poor neurological outcome at 6 months with 100% specificity and 20.8% sensitivity (very-low certainty of evidence).
In one study [Moon 2018 36, 44 pts] ADC volume proportion (400× 10−6 mm2/s) >2.5% at <48h predicted poor neurological outcome at 6 months with 100% specificity and 64% sensitivity (very-low certainty of evidence).
In one study [Moon 2018 36, 66 pts] ADC volume proportion (400× 10−6 mm2/s) >1.66% at 48h-7days predicted poor neurological outcome at 6 months with 100% specificity and 79.2% sensitivity (very-low certainty of evidence).
In one study [Kim 2013 1393, 51 pts] Maximum Cluster Size in different cerebral regions on CT ≤151.7× 10−6 mm2/s at 46 (37-52)h predicted poor neurological outcome at 6 months with 100% specificity and sensitivity ranging from 62.5% to 90% (very-low certainty of evidence).
In one study [Kim 2013 1393, 51 pts] the Lowest Mean ADC in different cerebral regions on CT ≤555.7× 10−6 mm2/s at 46 (37-52)h predicted poor neurological outcome at 6 months with 100% specificity and sensitivity ranging from 50% to 72.5% (very-low certainty of evidence).
In one study [Kim 2013 1393, 51 pts] the Lowest Minimum ADC in different cerebral regions on CT ≤466.8× 10−6 mm2/s at 46 (37-52)h predicted poor neurological outcome at 6 months with 100% specificity and sensitivity ranging from 42.5% to 82.5% (very-low certainty of evidence).
In one study [Hirsch 2019 in press, 51 pts] ≥10% of ADC<650× 10−6 mm2/s at 1-7 days predicted poor neurological outcome at 6 months with 100% specificity and 51.4% sensitivity (very-low certainty of evidence).
Treatment recommendations
- We suggest using grey matter/white matter (GWR) ratio on brain CT for predicting neurological outcome of adults who are comatose after cardiac arrest (weak recommendation, very-low-certainty evidence). However, no GWR threshold for 100% specificity can be recommended.
- We suggest using DWI on brain MRI for predicting neurological outcome of adults who are comatose after cardiac arrest (weak recommendation, very-low-certainty evidence).
- We suggest using ADC on brain MRI for predicting neurological outcome of adults who are comatose after cardiac arrest (weak recommendation, very-low-certainty evidence).
Justification and Evidence to Decision Framework Highlights
As for the 2015 CoSTR on this topic, the Task Force opinion is that a multimodal approach should be used in all cases with all supplementary tests considered in the context of prognostication.
Severe brain oedema in patients who are unconscious after cardiac arrest predicts poor outcome with high specificity. GWR allows a quantitative evaluation of brain oedema. However, there is a wide heterogeneity of measurement techniques (sites and calculation methods) for GWR. This may partly explain the wide variability of thresholds for 100% specificity across the studies we included. The evidence supporting GWR has very low certainty.
Assessing DWI has a potential for predicting poor neurological outcome after cardiac arrest. The definition of what a positive DWI MRI after cardiac arrest was inconsistent or even absent in the studies we included. The supporting evidence had very low certainty.
Assessing apparent diffusion coefficient (ADC) has a potential for predicting poor neurological outcome after cardiac arrest with high sensitivity. There is a wide heterogeneity of measurement techniques (sites and calculation methods) for ADC across studies. The supporting evidence for ADC had very low certainty.
Knowledge Gaps
A consistent GWR threshold for predicting poor neurological outcome after cardiac arrest should be identified.
A standardisation of the methods for GWR calculation is warranted.
The optimal timing for prognostication using brain CT after cardiac arrest is still unknown. Studies assessing serial brain CT after cardiac arrest are desirable.
The criteria for defining a positive DWI MRI after cardiac arrest need to be standardised
A consistent ADC threshold for predicting poor neurological outcome after cardiac arrest should be identified.
A standardisation of the methods for ADC calculation is warranted.
Attachments
Evidence-to-Decision Table: GWR ETD
Evidence-to-Decision Table: DWI ETD
Evidence-to-Decision Table: ADC ETD
References
Jang J, Oh SH, Nam Y, Lee K, Choi HS, Jung SL, Ahn KJ, Park KN, Kim BS. Prognostic value of phase information of 2d t2*-weighted gradient echo brain imaging in cardiac arrest survivors: A preliminary study. Resuscitation. 2019; 140:142-149
Jeon CH, Park JS, Lee JH, Kim H, Kim SC, Park KH, Yi KS, Kim SM, Youn CS, Kim YM, Lee BK. Comparison of brain computed tomography and diffusion-weighted magnetic resonance imaging to predict early neurologic outcome before target temperature management comatose cardiac arrest survivors. Resuscitation. 2017; 118:21-26
Kim J, Kim K, Hong S, Kwon B, Yun ID, Choi BS, Jung C, Lee JH, Jo YH, Kim T, Rhee JE, Lee SH. Low apparent diffusion coefficient cluster-based analysis of diffusion-weighted MRI for prognostication of out-of-hospital cardiac arrest survivors. Resuscitation. 2013; 84:1393-1399
Kim JH, Kim MJ, You JS, Lee HS, Park YS, Park I, Chung SP. Multimodal approach for neurologic prognostication of out-of-hospital cardiac arrest patients undergoing targeted temperature management. Resuscitation. 2018; 134:33-40
Kim Y, Ho Lee J, Kun Hong C, Won Cho K, Hoon Yeo J, Ju Kang M, Weon Kim Y, Yul Lee K, Joo Kim J, Youn Hwang S. Feasibility of optic nerve sheath diameter measured on initial brain computed tomography as an early neurologic outcome predictor after cardiac arrest. Academic emergency medicine : official journal of the Society for Academic Emergency Medicine. 2014; 21:1121-1128
Lee BK, Jeung KW, Lee HY, Jung YH, Lee DH. Combining brain computed tomography and serum neuron specific enolase improves the prognostic performance compared to either alone in comatose cardiac arrest survivors treated with therapeutic hypothermia. Resuscitation. 2013;84:1387-1392
Lee BK, Jeung KW, Song KH, Jung YH, Choi WJ, Kim SH, Youn CS, Cho IS, Lee DH. Prognostic values of gray matter to white matter ratios on early brain computed tomography in adult comatose patients after out-of-hospital cardiac arrest of cardiac etiology. Resuscitation. 2015; 96:46-52
Lee BK, Kim WY, Shin J, Oh JS, Wee JH, Cha KC, Park Y, Choi JH, Jeung KW. Prognostic value of gray matter to white matter ratio in hypoxic and non-hypoxic cardiac arrest with non-cardiac etiology. The American journal of emergency medicine. 2016; 34:1583-1588
Lee DH, Lee BK, Jeung KW, Jung YH, Cho YS, Cho IS, Youn CS, Kim JW, Park JS, Min YI. Relationship between ventricular characteristics on brain computed tomography and 6-month neurologic outcome in cardiac arrest survivors who underwent targeted temperature management. Resuscitation. 2018; 129:37-42
Lee KS, Lee SE, Choi JY, Gho YR, Chae MK, Park EJ, Choi MH, Hong JM. Useful computed tomography score for estimation of early neurologic outcome in post-cardiac arrest patients with therapeutic hypothermia. Circulation journal : official journal of the Japanese Circulation Society. 2017; 8:1628-1635
Moon HK, Jang J, Park KN, Kim SH, Lee BK, Oh SH, Jeung KW, Choi SP, Cho IS, Youn CS. Quantitative analysis of relative volume of low apparent diffusion coefficient value can predict neurologic outcome after cardiac arrest. Resuscitation. 2018; 126:36-42
Scarpino M, Lanzo G, Lolli F, Carrai R, Moretti M, Spalletti M, Cozzolino M, Peris A, Amantini A, Grippo A. Neurophysiological and neuroradiological multimodal approach for early poor outcome prediction after cardiac arrest. Resuscitation. 2018; 129:114-120
Wang GN, Chen XF, Lv JR, Sun NN, Xu XQ, Zhang JS. The prognostic value of gray-white matter ratio on brain computed tomography in adult comatose cardiac arrest survivors. Journal of the Chinese Medical Association: JCMA. 2018; 81:599-604
Youn CS, Callaway CW, Rittenberger JC. Combination of initial neurologic examination, quantitative brain imaging and electroencephalography to predict outcome after cardiac arrest. Resuscitation. 2017;110:120-125