Access the full text.
Sign up today, get DeepDyve free for 14 days.
Loading next page...
/lp/american-journal-of-neuroradiology/parieto-occipital-injury-on-diffusion-mri-correlates-with-poor-rS0lZhe9uM
References
References for this paper are not available at this time. We will be adding them shortly, thank you for your patience.
- Publisher
- American Journal of Neuroradiology
- Copyright
- © 2023 by American Journal of Neuroradiology. Indicates open access to non-subscribers at www.ajnr.org
- ISSN
- 0195-6108
- eISSN
- 1936-959X
- DOI
- 10.3174/ajnr.a7779
- Publisher site
- See Article on Publisher Site
Abstract
ORIGINAL RESEARCH ADULT BRAIN Parieto-Occipital Injury on Diffusion MRI Correlates with Poor Neurologic Outcome following Cardiac Arrest E. Calabrese, S. Gandhi, J. Shih, M. Otero, D. Randazzo, C. Hemphill, R. Huie, J.F. Talbott, and E. Amorim ABSTRACT BACKGROUND AND PURPOSE: MR imaging of the brain provides unbiased neuroanatomic evaluation of brain injury and is useful for neurologic prognostication following cardiac arrest. Regional analysis of diffusion imaging may provide additional prognostic value and help reveal the neuroanatomic underpinnings of coma recovery. The purpose of this study was to evaluate global, regional, and voxelwise differences in diffusion-weighted MR imaging signal in patients in a coma after cardiac arrest. MATERIALS AND METHODS: We retrospectively analyzed diffusion MR imaging data from 81 subjects who were comatose for .48 hours following cardiac arrest. Poor outcome was defined as the inability to follow simple commands at any point during hospitalization. ADC differences between groups were evaluated across the whole brain, locally by using voxelwise analysis and regionally by using ROI-based principal component analysis. RESULTS: Subjects with poor outcome had more severe brain injury as measured by lower average whole-brain ADC (740 [SD, 6 2 6 2 6 102] 10 mm /s versus 833 [SD, 23] 10 mm /s, P, .001) and larger average volumes of tissue with ADC below 650 10 mm /s (464 [SD, 469] mL versus 62 [SD, 51] mL, P, .001). Voxelwise analysis showed lower ADC in the bilateral parieto-occipital areas and perirolandic cortices for the poor outcome group. ROI-based principal component analysis showed an association between lower ADC in parieto-occipital regions and poor outcome. CONCLUSIONS: Brain injury affecting the parieto-occipital region measured with quantitative ADC analysis was associated with poor outcomes after cardiac arrest. These results suggest that injury to specific brain regions may influence coma recovery. ABBREVIATIONS: AUC ¼ area under the curve; CPC ¼ Cerebral Performance Category; PC ¼ principal component; PCA ¼ principal component analysis; PCAC ¼ post–cardiac arrest coma; ROC ¼ receiver operating characteristic; ROSC ¼ return of spontaneous circulation; TTM ¼ targeted temperature manage- ment; WLST ¼ withdrawal of life-sustaining therapies ardiac arrest affects 600,000 patients annually in the United For patients surviving initial resuscitation, mortality ranges from CStates despite substantial advances in resuscitation medicine. 75% to 88% for in-hospital and out-of-hospital cardiac arrests, respectively. Most deaths follow withdrawal of life-sustaining therapies (WLST) due to perceived poor neurologic prognosis after Received July 28, 2022; accepted after revision January 3, 2023. multimodal evaluation with serial examinations and ancillary test- From the Department of Radiology and Biomedical Imaging (E.C., S.G., J.F.T.), ing. Determination of prognosis is, therefore, a key component of Department of Neurology (J.S., M.O., D.R., C.H., E.A.), Weill Institute for Neurosciences, and Department of Neurological Surgery (R.H.), University of post–cardiac arrest care, and accurate evaluation of brain injury se- California, San Francisco, San Francisco, California; and Department of Radiology and verity is critical to prevent premature WLST in post–cardiac arrest Biomedical Imaging (S.G., J.F.T., E.A.), Zuckerberg San Francisco General Hospital, San Francisco, California. coma (PCAC) who have the potential to recover. This study was supported by the American Heart Association (20CDA35310297; The inclusion of MR imaging, particularly DWI, in the multi- 2020AMFDP), the National Institutes of Health (1K23NS119794), and the National modal neuroprognostication paradigm has gained traction in Center for Advancing Translational Sciences, National Institutes of Health, through the University of California, San Francisco–Clinical Translation and Science routine clinical care, particularly in the targeted temperature Institute, grant No. UL1 TR001872. management (TTM) era. The use of sedatives and neuromuscu- The contents of this study are solely the responsibility of the authors and do not lar blockade as well as metabolic changes with TTM may decrease necessarily represent the official views of the National Institutes of Health. the validity of previously established prognostication tools such as Please address correspondence to Edilberto Amorim, MD, Department of Neurology, Weill Institute for Neurosciences, University of California, San Francisco, Zuckerberg clinical examination, electroencephalography, and somatosensory- San Francisco General Hospital, 1001 Potrero Ave, Building 1, Suite 312, San Francisco, evoked potentials, while abnormalities on MR imaging are unlikely CA, 94110; e-mail: edilbertoamorim@gmail.com; @EdAmorimMD; @ecalabr to be affected. Research-leveraging qualitative and quantitative MR Indicates open access to non-subscribers at www.ajnr.org imaging analysis has demonstrated utility for outcome predictions Indicates article with online supplemental data. 3-16 http://dx.doi.org/10.3174/ajnr.A7779 in post–cardiac arrest coma. However, MR imaging literature 254 Calabrese Mar 2023 www.ajnr.org for outcome prediction after cardiac arrest has been limited by vari- follow commands were categorized as having good outcome and ability in MR imaging scanning methodology, qualitative rather patients unable to follow commands as having poor outcome. than quantitative assessment, and a focus on global rather than re- Image Acquisition gional patterns of brain injury. While several prior studies have MR images were acquired on a single 3T MR scanner (Magnetom demonstrated that the presence of any acute brain injury is associ- Skyra; Siemens). Imaging protocol included either a 6- or 12-direc- ated with poor outcome, relatively fewer studies have addressed re- tion DWI with b ¼ 1000 s/mm . ADC maps were generated auto- gional brain injury patterns and their association with good-versus- matically on the scanner. Additional DWI parameters included an 6,7,9,10,12-16 poor clinical outcome. FOV of 22 22 cm, an image matrix of 192 192, and section The main purpose of this study was to quantitatively describe thickness of 4 mm (voxel resolution ¼ 1.14 1.14 4 mm), sec- the regional neuroanatomic distribution of brain injury post- tion gap of 1 mm, section number range of 28–34, TR range of cardiac arrest using DWI. We found that both global brain injury 4710–5520 ms, TE range of 64–74 ms, and number of excitations and specifically injury to the parieto-occipital region were associ- of 1. Representative examples of DWI and ADC images are pro- ated with severe neurologic impairment. These results provide vided in the Online Supplemental Data. important insights into the clinical and prognostic significance of regional brain injury patterns on diffusion-weighted MR imaging Image Preprocessing for post–cardiac arrest coma and address existing knowledge gaps DWI volumes were processed using custom software pro- in the importance of assessment of global-versus-regional brain grammed in Python 3.8 using the Nipype 1.6 package (https:// injury. github.com/nipy/nipype/releases/tag/1.6.0) to interface with non- Python software. DWI volumes were converted to NIfTI format using dcm2niix 1.0 (https://github.com/rordenlab/dcm2niix/ MATERIALS AND METHODS releases) and then aligned to the Montreal Neurological Institute Patient Selection 152 brain atlas using multistep rigid, affine, and diffeomorphic This study was approved by the University of California San automated image registration implemented using Advanced Francisco and Zuckerberg San Francisco General Hospital institu- Normalization Tools (ANTs 2.3.5; http://stnava.github.io/ANTs/). tional review boards with a waiver of informed consent. Using elec- Qualitative registration accuracy was manually evaluated for qual- tronic health record search tools, we identified all subjects treated at ity control. Resulting transforms were used to map patient ADC a single university-affiliated hospital between 2016 and 2021 for maps into the atlas space. Representative examples of DWI and post–cardiac arrest coma (n ¼ 335). For this study, post–cardiac ADC images before and after registration are included in the arrest coma were defined as patients having a Glasgow Coma Scale Online Supplemental Data. score of #8 after return of spontaneous circulation (ROSC). We then identified patients who underwent inpatient brain MR imaging ROI Extraction (n ¼ 105) as part of a routine multimodal prognostic evaluation for A total of 124 bilateral cortical (n ¼ 96) and subcortical (n ¼ 28) patients who do not recover consciousness after the rewarming ROIs were extracted from each patient’sDWI volume using the phase of TTM or by 48 hours following arrest. Exclusion criteria Harvard-Oxford cortical and subcortical brain atlases, respectively. included patients with missing MR imaging sequences (n ¼ 1), All structures were assessed in both the left and right hemispheres patients who did not complete TTM (n ¼ 2), the presence of a sig- individually, and subcortical structures were also evaluated as a nificant unrelated abnormality on brain MR imaging such as hem- single bilateral ROI. orrhage or large-territory encephalomalacia (n ¼ 8), or MR imaging Quantitative ADC Analysis acquired.7days (168hours) post–cardiac arrest. Poor outcome Brain-wide ADC differences between outcome groups were was defined as the inability to follow 1-step commands before hos- assessed using 2 metrics established in previously published work: pital discharge on the basis of retrospective review of the electronic average whole-brain ADC and total volume of tissue with ADC health record. 6 s 8 less than 650 10 mm /s (ADC ). Voxels with ADC values Clinical notes by the primary team, neurology, occupational 6 2 greater than 1100 10 mm /s were considered to represent therapy, and/or physical therapy were reviewed to determine the CSF and were excluded. Regional mean ADC and ADC differ- best clinical examination before discharge. The ability to follow ences between groups were also assessed for each individual region commands was part of the standard evaluation from clinical in the Harvard-Oxford cortical and subcortical brain atlases. providers, which includes evaluation of at least 1-step axial and appendicular commands (eg, show fingers, wiggle toes, or stick Whole-Brain Injury Frequency Mapping out tongue). Cerebral Performance Category (CPC) scores were Whole-brain injury frequency maps were generated to allow visu- recorded for all patients included in the analysis but were assessed alization of common sites of injury for each outcome group. Brain only at the time of discharge and, therefore, did not capture injury was defined as all brain voxels with ADC less than 650 6 s patients who recovered consciousness and the ability to follow 10 mm /s. Binarized injury maps were generated for each commands during hospitalization but subsequently died after patient and averaged within each outcome group. The output of WLST (n ¼ 3 in this cohort). With the exception of these 3 this analysis was a single injury-frequency map for each outcome patients, the good and poor outcome groups corresponded group, in which voxel values represent the proportion of patients exactly to CPC 1–3and CPC 4–5, respectively. Patients able to in the outcome group with injury in that location. AJNR Am J Neuroradiol 44:254–60 Mar 2023 www.ajnr.org 255 Subject demographics stratified by outcome group construct of each PC. PC scores for Follows Commands Does Not Follow each patient were then used for hy- Variable (n = 17) Commands (n = 64) P Value pothesis testing. A backward regres- Age (mean) (yr) 55 (SD, 16) 57 (SD, 15) .741 sion that included PC scores and other Female sex 4/17 (24%) 18/64 (28%) .701 a clinical variables was used to deter- Race and ethnicity .757 mine significant predictors of poor Asian 4 (21%) 6 (9%) Black 3 (16%) 14 (22%) outcome. The resulting significant pre- Hispanic 2 (11%) 6 (9%) dictors were included in a receiver Hawaiian 0 (0%) 2 (3%) operating characteristic (ROC) curve Non-Hispanic white 2 (11%) 6 (9%) analysis to determine the sensitivity Unknown 8 (42%) 30 (47%) and specificity of these variables to pre- Time from ROSC to MR 120 (SD, 26) 115 (SD, 27) .496 imaging (mean) (hr) dict poor outcome. PCA and ROC ana- Shockable rhythm 6/17 (35%) 12/64 (19%) .145 lyses were run using the syndRomics Out-of-hospital cardiac arrest 14/17 (82%) 49/64 (77%) .610 (https://github.com/ucsf-ferguson-lab/ Time to ROSC (mean) 19 (SD, 14) 24 (SD, 14) .182 syndRomics) and pROC software pack- Patient race and ethnicity do not add to 100% as some patients reported more than one race and ethnicity ages in the R statistical framework (R combination. Version 4.1.2; http://www.r-project. org/). Statistical Analysis Continuous variables and categoric variables are reported as means (SD) and frequencies with percentages. The x test was used to compare categoric data. The 2-sample Student t test was used to com- pare continuous variables. The DeLong method was used to compare ROC curves. For all statistical tests, a P value , .05 was considered statistically significant. For the ROI analysis, a Bonferroni correction was used to control for multiple comparisons. RESULTS Subjects We identified a total of 335 subjects: Two hundred fifty-four FIG 1. Frequency of hypoxic-ischemic brain injury in both outcome were excluded, and 81 were included in the analysis (Online groups. Colorized 3D brain renderings show the frequency of injury Supplemental Data). Sixty-four of 81 (79%) subjects had poor (as defined by ADC, 650 10–6mm /s) for the poor-outcome group (A) and the good outcome group (B), respectively. The color outcomes. Demographic information for each outcome group is bar (right) indicates the frequency of injury across the whole brain. presented in the Table. There were no statistically significant dif- Note that regions with,5% injury frequency are transparent to allow ferences in age, sex, race, and ethnicity; time to MR imaging; the better visualization of more frequently injured areas. presence of a shockable rhythm at presentation; location of car- diac arrest (in-hospital versus out-of-hospital); or time to ROSC. Voxelwise Quantitative ADC Analysis When we compared patients who were included versus excluded Voxelwise ADC differences between outcome groups were from the analysis, there was no statistically significant difference assessed using threshold-free cluster enhancement implemented in in the proportion of patients who had WLST or brain death (x FSL 6.0.2 (http://www.fmrib.ox.ac.uk/fsl). Statistical significance P value ¼ .12 and .89, respectively) (Online Supplemental Data). was determined as a family-wise error–corrected P value, .05. Several potentially confounding variables were included as con- Quantitative ADC Comparison between Outcome Groups found regressors in the analysis: whole-brain average ADC, patient The poor outcome group had lower whole-brain mean ADC val- 6 2 6 age, patient sex, and time to MR imaging (in hours). ues (740 [SD, 102] 10 mm /s versus 833 [SD, 23] 10 mm /s, P, .001) and larger total ADC volumes (464 [SD, ROI-Based Principal Component Analysis 469] mL versus 62 [SD, 51] mL, P, .001). Statistically significant Principal component analysis (PCA) was used to explore rela- differences in mean ADC and ADC by ROI (after Bonferroni tionships between ROI-based injury and clinical outcome. The correction) are presented in the Online Supplemental Data. measure used for PCA was the percentage of brain parenchyma 6 s with ADC less than 650 10 mm /s for each ROI. A scree Whole-Brain Injury Mapping plot was generated to determine the percentage of variance Whole-brain ADC frequency maps were generated for each accounted for by each principal component (PC). Variable load- outcome group and are presented in Fig 1. The poor outcome ings within each PC were thresholded at an absolute value of group had a more severe and extensive pattern of brain injury 0.5, and expert visual analysis of loading patterns for these most frequently involving the precentral gyri, precuneus, and strong loading variables was used to identify the underlying cuneus (Fig 1A). The good outcome group (Fig 1B)showed a 256 Calabrese Mar 2023 www.ajnr.org relatively heterogeneous pattern of injury, with the most frequent PCA of ROI-Based ADC Differences between Outcome areas of injury being the cerebellum, globi pallidi, and precentral Groups gyri. PCA analysis of the percentage of brain parenchyma with ADC 6 s less than 650 10 mm /s for each ROI identified 2 PCs that Voxelwise Analysis of Quantitative ADC Differences accounted for .95% of the variance. PC loadings for PC1 and between Outcome Groups PC2 are included as Online Supplemental Data. All ROI-based Voxelwise analysis with threshold-free cluster enhancement variables loaded on PC1 in the same direction, indicating that was used to explore regional differences in ADC values between they all contributed to the variance explained by this PC (79%). outcome groups without a priori anatomic constraints. Results In contrast, ROI-based variables had variable loading contribu- of the voxelwise analysis are seen in Fig 2. Statistically signifi- tions to the variance explained by PC2 (16%), with some variables cant clusters (after controlling for whole-brain average ADC, (perirolandic and parieto-occipital ROIs) showing a positive cor- patient age, patient sex, and time from ROSC to MR imaging relation and others (anterior-frontotemporal and deep gray matter as confounding variables) were localized to the cuneus and ROIs) showing a negative correlation (Fig 3). precuneus. ROC Analysis of Outcome Groups with PCA Results ROC analysis for determining neurologic outcome on the basis of clinical and MR imaging features is presented in Fig 4.Using only the percentage of the whole brain with ADC less than 650 10 mm /s yielded a relatively high discriminative performance, with an ROC area under the curve (AUC) of 0.84. Whole-brain injury of 10% was 95% specific and 59% sensitive for poor outcome pre- diction, and whole-brain injury of 5% was 74% specific and 77% sensitive. ROC performance was similar using PC2 scores with an ROC AUC of 0.83. By comparison, the single best ROI-based pre- dictor (average percentage of the cuneal cortex with ADC less 6 s than 650 10 mm /s) yielded a similar performance with an ROC AUC of 0.89. An average cuneal injury of 5% was 94% spe- cific and 75% sensitive. Combining PC scores with the location of arrest (in-hospital versus out-of-hospital) yielded the best per- formance with an ROC AUC of 0.91. Pair-wise comparison of ROC curves using the DeLong method revealed a statistically sig- nificant difference between the PC2 and PC11 PC21 arrest loca- tion curves (P ¼ .029), with all other comparisons failing to reach statistical significance (Online Supplemental Data). ROC analysis using CPC scores rather than the ability to follow commands as the outcome metric is included as Online Supplemental Data. FIG 2. Results of voxelwise comparisons between outcome groups. Statistically significant voxels (family-wise error rate–corrected, P, .05) are shown as color overlays on orthogonal slices of the DISCUSSION Montreal Neurologic Institute brain atlas (A–C). Color indicates the Regional analysis of quantitative ADC showed that involvement P value for the comparison (see the color bar). A volume-rendering of the cuneus, precuneus, and precentral gyrus was associated of the P value from a lateral perspective is also shown, with dashed with poor outcomes post–cardiac arrest. Most important, whole- yellow lines indicating the plane of axial and coronal slices (D). brain ADC frequency mapping revealed that injury to the precentral gyrus was more prominent in patients with poor outcome, but could also be present in patients with good outcome. However, injury to the parieto-occipital region was more specifically associated with poor outcomes as highlighted by both the PCA and voxelwise analyses, includ- ing when controlling for whole-brain ADC, which is expected to remove some of the variance explained by the FIG 3. Loading values for PC2 from the PCA. Each cortical region from the Harvard Oxford brain severity of global brain injury. The per- atlas is colored according to its PC2 loading value. Red and blue loadings indicate opposite pat- centage of ADC in the cuneus alone terns of loading (red indicates positive correlation, blue indicates a negative correlation). The showed high performance for determin- dashed yellow line on the lateral projection (right) indicates the plane of the axial section (center). ing neurologic outcome, with a value of AJNR Am J Neuroradiol 44:254–60 Mar 2023 www.ajnr.org 257 FIG 4. ROC analysis of different diffusion-weighted MR imaging–based variables for predicting outcome in patients with post–cardiac arrest coma. OHCA indicates out of hospital cardiac arrest. 5%, yielding 75% sensitivity and 94% specificity for poor outcome. Injury to several other specific brain regions has also been associ- These findings corroborate previous studies demonstrating that ated with poor outcome; however, most studies pursued a descrip- injury to this region is associated with poor outcomes in post– tive approach that did not incorporate multivariable analytic cardiac arrest coma; however, there has been limited prior work models, which may help determine which regions are specifically directly investigating regional injury correlates for poor clinical useful for discriminating between good and poor outcomes inde- outcome in cardiac arrest and none using the PCA and voxelwise pendent of diffuse brain injury. 9,10,15,18 analyses presented here. Additional investigation is needed The regional vulnerability of the precuneus, cuneus lobes, and to determine whether injury to the parieto-occipital region is precentral gyrus following cardiac arrest is not well-understood. exclusively a marker of severe injury or a causal determinant of The precuneus, cuneus, visual cortex, and precentral gyrus are persistent coma post–cardiac arrest. considered some of the most metabolically active regions in The posterior-medial cortex is part of the default mode net- humans; therefore, this finding could be a consequence of meta- work and is involved in regulation of consciousness and connec- bolic demand mismatch in the context of hypoxic-ischemic brain 19,20 26 tivity throughout the brain. Resting-state fMRI studies injury and reperfusion injury following cardiac arrest. An alter- involving patients with both severe brain injury of distinct etiolo- native vascular hypothesis is that the posterior circulation cannot gies and various levels of consciousness showed that the degree of provide sufficient cerebral blood flow during cardiopulmonary connectivity of the default mode network, in particular of the pre- resuscitation and its limited autoregulatory response following cuneus or posterior cingulate cortex, was directly proportional to return of spontaneous circulation and reperfusion may contrib- the level of consciousness, with patients in a coma having lower ute to the increased vulnerability of these regions, similar to what 19,21-24 27 connectivity. Our report adds to the post–cardiac arrest is observed in posterior reversible encephalopathy syndrome. coma literature demonstrating that lower ADC values in the pari- The use of quantitative MR imaging can augment multimodal 9,11,12,15,20,25 eto-occipital region is associated with poor outcomes. prognostication, and our study demonstrates the value of combining 258 Calabrese Mar 2023 www.ajnr.org clinical and neuroimaging information for outcome prediction. therefore, brain injury burden likely contributed to self-fulfilling Several different clinical, laboratory, and electrophysiologic parame- prophecies related to WLST due to perceived poor neurologic 9,12,13 ters are known to be associated with clinical outcomes in patients prognosis. However, we did not detect any significant differ- with cardiac arrest; however, most studies focused on imaging- ence in the rates of WLST or brain death between patients who based predictors have favored a global rather than region-specific were included versus excluded in the analysis. Finally, the rela- analysis. We showed that outcome determination using individ- tively small sample size and small proportion of patients with ual regions, whole brain, or a combination of specific regions using good outcome limit our ability to generalize findings or draw con- PCA had excellent performance, with AUCs ranging from 0.84 to clusions about when WLST should be considered. Larger and pro- 0.91. Our results are consistent with those in previous literature spective studies focused on assessment of regional brain injury showing that the mean ADC as well as regional ADC are helpful will be needed to confirm the association of parieto-occipital 13-15 for prognostication in post–cardiac arrest coma. The specific injury with poor clinical outcome. cutoff for ADC has not yet been defined, with a cutoff value of .10%–26% brain involvement being needed to achieve specificity CONCLUSIONS 4,5,8,14 for poor outcome prediction (ie, above 95%). In our study, Injury to the parieto-occipital region is associated with poor out- 10% brain involvement with ADC was 95% specific and 55% comes following cardiac arrest. Involvement of the precentral sensitive for predicting poor outcome, and 5% brain involvement gyrus can be seen in both good and poor outcome groups; how- was 74% specific and 77% sensitive. In addition, previous studies ever, patients recovering from coma had a much lower burden of have shown that lower thresholds for ADC for such as 400 or 450 brain injury. Regional and voxelwise analysis of quantitative ADC 6 2 10 mm /s yield higher ORs for predicting poor neurologic may add value to multimodal neurologic prognostication in com- 6,8,16 outcomes. Additional studies exploring the ADC parameter bination with other established predictors and provide informa- space using larger and more racially and ethnically diverse cohorts tion about the neuroanatomic regions determining the potential may help define global and regional thresholds with the best per- for coma recovery. formance for outcome prediction. Some of the variability in previously published data on the Disclosure forms provided by the authors are available with the full text and PDF of this article at www.ajnr.org. predictive value of DWI in patients with cardiac arrest could be secondary to differences in patient selection, variable timing of REFERENCES MR imaging, variable methodology, and differences in outcome 1. Virani SS,Alonso A,Aparicio HJ,et al;AmericanHeart Association selection. Given these limitations, in addition to limited access to Council on Epidemiology and Prevention Statistics Committee and quantitative DWI analysis techniques, we believe the results from Stroke Statistics Subcommittee. Heart Disease and Stroke Statistics-2021 regional analysis may be add important insight for clinical prac- Update: a report from the American Heart Association. Circulation tice. Temporal variations in DWI appearance in patients after 2021;143:e254–743 CrossRef Medline cardiac arrest have been established by multiple observational 2. Madden LK, Hill M, May TL, et al. The implementation of targeted temperature management: an evidence-based guideline from the and quantitative studies, with the ability to predict outcome being Neurocritical Care Society. Neurocrit Care 2017;27:468–87 CrossRef 6,13,15 best on MR imaging performed on days 3–5. These observa- Medline tions likely reflect both the well-described phenomenon of infarct 3. An C, You Y, Park JS, et al. Thecut-off valueof a qualitativebrain dif- pseudonormalization on DWI and the complex pathophysiology fusion-weighted image (DWI) scoring system to predict poor neuro- of anoxic injury, with a significant component of secondary logic outcome in out-of-hospital cardiac arrest (OHCA) patients after target temperature management. Resuscitation 2020;157:202–10 injury caused by oxidative stress and neuronal excitotoxicity. CrossRef Medline For these reasons, we excluded from our analysis patients with 4. Wouters A, Scheldeman L, Plessers S, et al. Added value of quantita- MR imaging performed .7 days following arrest; however, it tive apparent diffusion coefficient values for neuroprognostication remains unclear whether more stringent time-to–MR imaging after cardiac arrest. Neurology 2021;96:e2611–18 CrossRef Medline limitations should be imposed for this type of study. Our study 5. Bevers MB, Scirica BM, Avery KR, et al. Combination of clinical exam, MRI and EEG to predict outcome following cardiac arrest benefited from imaging acquisition in a single scanner with com- and targeted temperature management. Neurocrit Care 2018;29:396– parable protocols, which may have minimized some of the scan- 403 CrossRef Medline ning-variability limitations. 6. Reynolds AS, Guo X, Matthews E, et al. Post-anoxic quantitative MRI There are several limitations of our study. This was a retro- changes may predict emergence from coma and functional out- spective and single-center study that relied on documentation of a comes at discharge. Resuscitation 2017;117:87–90 CrossRef Medline 7. Kim J, Kim K, Suh GJ, et al. Prognostication of cardiac arrest survi- clinical examination documentation for outcome determination. vors using low apparent diffusion coefficient cluster volume. While our outcome metric closely followed discharge CPC scores, Resuscitation 2016;100:18–24 CrossRef Medline it is, nonetheless, limited due to lack of standardization. In addi- 8. Hirsch KG, Mlynash M, Eyngorn I, et al. Multi-center study of diffu- tion, at our institution the use of brain MR imaging is part of a sion-weighted imaging in coma after cardiac arrest. Neurocrit Care multimodal prognostication protocol, together with other ancil- 2016;24:82–89 CrossRef Medline 9. Kim J, Kim K, Hong S, et al. Low apparent diffusion coefficient lary tests for patients who do not recover consciousness after cluster-based analysis of diffusion-weighted MRI for prognostica- TTM, possibly leading to a selection bias toward evaluating tion of out-of-hospital cardiac arrest survivors. Resuscitation patients with more severe brain injury. While quantitative analysis 2013;84:1393–99 CrossRef Medline of ADC data was not available for decision-making, clinicians 10. Kim J, Choi BS, Kim K, et al. Prognostic performance of diffusion- were not blinded to qualitative MR imaging interpretation; weighted MRI combined with NSE in comatose cardiac arrest AJNR Am J Neuroradiol 44:254–60 Mar 2023 www.ajnr.org 259 survivors treated with mild hypothermia. Neurocrit Care 2012; 20. Li R, Utevsky AV, Huettel SA, et al. Developmental maturation of 17:412–20 CrossRef Medline the precuneus as a functional core of the default-mode network. 2019;31:1506–19 CrossRef Medline 11. Choi SP, Park KN, Park HK, et al. Diffusion-weighted magnetic res- 21. Silva S, de Pasquale F, Vuillaume C, et al. Disruption of posteromedial onance imaging for predicting the clinical outcome of comatose large-scale neural communication predicts recovery from coma. survivors after cardiac arrest: a cohort study. Crit Care 2010;14:R17 Neurology 2015;85:2036–44 CrossRef Medline CrossRef Medline 22. Koenig MA, Holt JL, Ernst T, et al. MRI default mode network con- 12. Hirsch KG, Mlynash M, Jansen S, et al. Prognostic value of a qualita- nectivity is associated with functional outcome after cardiopulmo- tive brain MRI scoring system after cardiac arrest. J Neuroimaging nary arrest. Neurocrit Care 2014;20:348–57 CrossRef Medline 2015;25:430–37 CrossRef Medline 23. Sair HI, Hannawi Y, Li S, et al; Neuroimaging for Coma Emergence 13. Wijman CAC, Mlynash M, Caulfield AF, et al. Prognostic value of and Recovery (NICER) Consortium. Early functional connectome brain diffusion-weighted imaging after cardiac arrest. Ann Neurol integrity and 1-year recovery in comatose survivors of cardiac 2009;65:394–402 CrossRef Medline arrest. Radiology 2018;287:247–55 CrossRef Medline 14. Hirsch KG, Fischbein N, Mlynash M, et al. Prognostic value of diffu- 24. Wagner F,Hänggi M,Weck A,et al. Outcome prediction with resting- sion-weighted MRI for post–cardiac arrest coma. Neurology 2020;94: state functional connectivity after cardiac arrest. Sci Rep 2020;10:11695 e1684–92 CrossRef Medline CrossRef Medline 15. Wu O, Sorensen AG, Benner T, et al. Comatose patients with car- 25. Ryoo SM, Jeon S-B, Sohn CH, et al; Korean Hypothermia Network diac arrest: predicting clinical outcome with diffusion-weighted Investigators. Predicting outcome with diffusion-weighted imaging in cardiac arrest patients receiving hypothermia therapy: multicen- MR imaging. Radiology 2009;252:173–81 CrossRef Medline ter retrospective cohort study. Crit Care Med 2015;43:2370–77 16. Moon HK, Jang J, Park KN, et al. Quantitative analysis of relative CrossRef Medline volume of low apparent diffusion coefficient value can predict neu- 26. Horwitz B, Duara R, Rapoport SI. Intercorrelations of glucose meta- rologic outcome after cardiac arrest. Resuscitation 2018;126:36–42 bolic rates between brain regions: application to healthy males in a CrossRef Medline state of reduced sensory input. J Cereb Blood Flow Metab 1984;4:484– 17. Torres-Espín A, Chou A, Huie JR, et al. Reproducible analysis of dis- 99 CrossRef Medline ease space via principal components using the novel R package 27. Anderson RC, Patel V, Sheikh-Bahaei N, et al. Posterior reversible syndRomics. eLife 2021;10:e61812 CrossRef Medline encephalopathy syndrome (PRES): pathophysiology and neuro- 18. Snider SB, Fischer D, McKeown ME, et al. Regional distribution of imaging. Front Neurol 2020;11:463 CrossRef Medline brain injury after cardiac arrest: clinical and electrographic corre- 28. Sandroni C, D’Arrigo S, Cacciola S, et al. Prediction of good neuro- lates. Neurology 2022;98:e1238–47 CrossRef Medline logical outcome in comatose survivors of cardiac arrest: a system- 19. Vanhaudenhuyse A, Noirhomme Q, Tshibanda LJ, et al. Default atic review. Intensive Care Med 2022;48:389–413 CrossRef Medline network connectivity reflects the level of consciousness in non- 29. Sekhon MS, AinsliePN, GriesdaleDE. Clinical pathophysiology of communicative brain-damaged patients. Brain 2010;133:161–71 hypoxic ischemic brain injury after cardiac arrest: a “two-hit” model. CrossRef Medline Crit Care 2017;21:90 CrossRef Medline 260 Calabrese Mar 2023 www.ajnr.org
Journal
American Journal of Neuroradiology – American Journal of Neuroradiology
Published: Mar 1, 2023