AAN.com
Quick search in Citations
AnywhereCitation
Advanced Search
Trending Terms:
×
Site maintenance Tuesday, June 24, 2025. Please note that new registrations and purchases will be unavailable on this date.
Research Article
May 30, 2024
Free Accessvideo

Dynamic Evolution of Infarct Volumes at MRI in Ischemic Stroke Due to Large Vessel Occlusion

This article has been corrected.
VIEW CORRECTION
Fanny Munsch, PhD https://orcid.org/0000-0002-7144-8617, David Planes, MD, Hikaru Fukutomi, MD, PhD, Gaultier Marnat, MD https://orcid.org/0000-0002-7611-7753, Thomas Courret, MD https://orcid.org/0000-0002-7282-1153, Emilien Micard, PhD, Bailiang Chen, PhD https://orcid.org/0000-0003-1038-7846, … Show All … , Pierre Seners, MD, PhD https://orcid.org/0000-0002-2134-0691, Johanna Dubos, MSc, Vincent Planche, MD, PhD https://orcid.org/0000-0003-3713-227X, Pierrick Coupé, PhD https://orcid.org/0000-0003-2709-3350, Vincent Dousset, MD, PhD https://orcid.org/0000-0003-0427-3236, Bertrand Lapergue, MD, PhD https://orcid.org/0000-0002-8993-2175, Jean Marc Olivot, MD, PhD https://orcid.org/0000-0003-2027-2276, Igor Sibon, MD, PhD https://orcid.org/0000-0002-1171-4215, Michel Thiebaut De Schotten, PhD https://orcid.org/0000-0002-0329-1814, and Thomas Tourdias, MD, PhD https://orcid.org/0000-0002-7151-6325 for the FRAME and ETIS investigatorsAuthors Info & Affiliations
June 25, 2024 issue
102 (12)
https://doi.org/10.1212/WNL.0000000000209427
Letters to the Editor
Track CitationsAdd to favorites
Short Formfull text

Abstract

Background and Objectives

The typical infarct volume trajectories in stroke patients, categorized as slow or fast progressors, remain largely unknown. This study aimed to reveal the characteristic spatiotemporal evolutions of infarct volumes caused by large vessel occlusion (LVO) and show that such growth charts help anticipate clinical outcomes.

Methods

We conducted a secondary analysis from prospectively collected databases (FRAME, 2017–2019; ETIS, 2015–2022). We selected acute MRI data from anterior LVO stroke patients with witnessed onset, which were divided into training and independent validation datasets. In the training dataset, using Gaussian mixture analysis, we classified the patients into 3 growth groups based on their rate of infarct growth (diffusion volume/time-to-imaging). Subsequently, we extrapolated pseudo-longitudinal models of infarct growth for each group and generated sequential frequency maps to highlight the spatial distribution of infarct growth. We used these charts to attribute a growth group to the independent patients from the validation dataset. We compared their 3-month modified Rankin scale (mRS) with the predicted values based on a multivariable regression model from the training dataset that used growth group as an independent variable.

Results

We included 804 patients (median age 73.0 years [interquartile range 61.2–82.0 years]; 409 men). The training dataset revealed nonsupervised clustering into 11% (74/703) slow, 62% (437/703) intermediate, and 27% (192/703) fast progressors. Infarct volume evolutions were best fitted with a linear (r = 0.809; p < 0.001), cubic (r = 0.471; p < 0.001), and power (r = 0.63; p < 0.001) function for the slow, intermediate, and fast progressors, respectively. Notably, the deep nuclei and insular cortex were rapidly affected in the intermediate and fast groups with further cortical involvement in the fast group. The variable growth group significantly predicted the 3-month mRS (multivariate odds ratio 0.51; 95% CI 0.37–0.72, p < 0.0001) in the training dataset, yielding a mean area under the receiver operating characteristic curve of 0.78 (95% CI 0.66–0.88) in the independent validation dataset.

Discussion

We revealed spatiotemporal archetype dynamic evolutions following LVO stroke according to 3 growth phenotypes called slow, intermediate, and fast progressors, providing insight into anticipating clinical outcome. We expect this could help in designing neuroprotective trials aiming at modulating infarct growth before EVT.

Introduction

In the event of large vessel occlusion stroke (LVOS), a cascade of cellular changes arises rapidly.1 This cascade begins with reversible electrical dysfunction (i.e., ischemic penumbra) and progresses to ion pumping and energy metabolism failures inducing growth of the irreversible infarct core if cerebral blood flow is not restored.1 The infarct growth rate (IGR) varies substantially from one patient to another,2 influenced by various factors—most notably, the ability of leptomeningeal collateral circulation to maintain residual perfusion, thereby minimizing the expansion of the infarct core into the penumbra.3,4 Endovascular thrombectomy (EVT) is the standard-of-care treatment to reopen anterior LVO and stop the ischemic cascade.5 The rate of infarct progression before EVT, which is now referred to as slow—intermediate—or fast progressors,6,7 represents the interindividual tolerance to ischemia and strongly correlates with clinical outcome after EVT.8 It is therefore crucial to improve our understanding on the growth phenotype before EVT, which will be one of the future therapeutic targets to continue improving the patients. Indeed, the growth phenotype could guide adjunctive therapy.9 The future challenge will be to interfere with stroke progression before EVT, which will require cerebroprotectant trials that will necessarily have to tailor specific neuroprotective strategy to specific growth phenotypes.
However, a major limitation is that there are currently no standard criteria to define slow, intermediate, or fast progressors, leaving us reliant on somewhat arbitrary definitions and thresholds.8,10-13 For example, fast progressors have been arbitrarily characterized by an infarct volume higher than 70 mL within the first 6 hours7,13 or an IGR higher than 5 mL/h12 or 10 mL/h8 among other definitions. One reason is the absence of comprehensive understanding of the archetypical dynamic evolutions of infarct volumes for the different growth phenotypes. The major challenge is that revealing the natural development of the infarct is impossible with a standard longitudinal approach because most patients are scanned a single time before attempting rapid recanalization whenever possible. Only few data reported 2 imaging sessions before recanalization.14 An alternative method would consist of inferring pseudo-longitudinal profiles from the concatenation of a large number of homogeneous cross-sectional data collected before recanalization akin to strategies used recently to map the typical evolution of brain volume across the lifespan15,16 or in neurodegenerative diseases.17,18 Indeed, if we could objectively attribute stroke patients presenting the same clot location to a specific growth phenotype, then, we could reconstruct stereotypical growth curves from such homogeneous cluster of patients explored at different time points after the onset of symptoms.
Based on these considerations, we aimed to identify slow, intermediate, and fast progressors following LVOS using nonsupervised clustering approach to reveal the characteristic spatiotemporal evolutions of infarct volumes according to such phenotypes of tolerance to ischemia. In addition, we seek to provide the first evidence supporting the role of these new charts in anticipating patients' outcomes.

Methods

Standard Protocol Approvals, Registrations, and Patient Consents

We conducted a secondary analysis using 2 prospective studies approved by the institutional ethics committee, the primary objectives of which were unrelated to this article. These are the French acute multimodal imaging to select patients for mechanical thrombectomy (FRAME; NCT03045146) and the Endovascular treatment in Ischemic stroke registry (ETIS; NCT03776877). Written informed consent was obtained from all patients or their legal representatives. Our analysis called DEVOL (Dynamic EVOLution of stroke) was reported following the Strengthening the Reporting of Observational Studies in Epidemiology criteria for observational studies.19

Patients

The FRAME prospectively recruited consecutive participants in 2 comprehensive stroke centers in France from January 2017 to February 2019 who presented with LVOS and underwent acute imaging including perfusion-weighted imaging (PWI) on arrival and were subsequently treated with EVT.20 Analysis of the images from FRAME patients has already been reported but focusing on penumbra20,21 while we explored the IGR in this study.
The ETIS is an ongoing, multicenter, prospective, real-life observational study evaluating LVOS treated with EVT in 21 comprehensive stroke centers in France. Images are now sent by the investigative centers and stored centrally (ETIS image subproject). There is a manual step of verification of the images and proper identification before their upload in a dedicated environment (ArchiMed). We considered the images available in ArchiMed for patients explored between January 2015 and January 2022. Details regarding data collection and materials have been previously published.22 Analysis from the images of ETIS patients has not been reported yet.
Within FRAME and ETIS databases, we selected the patients with:
1.
Anterior circulation LVO (intracranial internal carotid artery [ICA], M1 segment of the middle cerebral artery, tandem with proximal intracranial occlusions).
2.
Witnessed stroke onset.
3.
MRI as baseline imaging before any recanalization strategy.
We identified and excluded any redundant records, thereby ensuring that each patient included in DEVOL analysis was unique and not counted more than once, even if they appeared in both databases. We excluded patients with stroke-on-awakening or unprecise onset and those explored with computed tomography or whose MRI was of insufficient quality (Figure 1).
Figure 1 Flowchart
Inclusion and exclusion criteria for patients from FRAME and ETIS. DWI = diffusion-weighted imaging; ICA = internal carotid artery.
We accessed a first set of available data of 703 patients from the FRAME and ETIS cohorts used as a training dataset to model temporal, spatial, and clinical evolution. Later on, we got access to 101 additional new patients from the ETIS dataset to build an independent dataset for validation (Figure 1).

Image Analyses

All the images were analyzed centrally blinded to clinical data. eFigure 1 provides on overview on the image-derived analyses described further. The infarct core masks, used to measure infarct volume and calculate IGR in our study, were delineated from acute diffusion MRI using automatic detection from Olea Sphere (version SP34) based on an apparent diffusion coefficient (ADC) threshold of 620 × 10−6 mm2/s and manual correction relying on b1000 and ADC images afterward if needed. Then, we registered all the images and the associated core masks to the standard Montreal Neurologic Institute 152 (MNI152) template, as previously described.21 Although diffusion MRI scans have been collected without harmonization across centers (eTable 1), we combined them based on the robustness of this sequence to field strength and acquisition parameters.23
PWI from FRAME had already been reported elsewhere,20,21 and we processed the additional PWI available from ETIS with Olea Sphere (version SP34). The masks of time-to-maximum (Tmax) greater than 6 seconds and 10 seconds were automatically extracted, visually checked, and normalized to the MNI152 template.
We derived collateral status from the hypoperfusion intensity ratio (HIR), that is, the percentage of the perfusion lesion with severely delayed contrast arrival times according to the formula: HIR = volume of Tmax >10 seconds/Tmax >6 seconds. HIR was repeatedly identified as a reliable proxy of good collaterals defined by angiography when HIR <0.4.24-26 When PWI was unavailable, we used data from the conventional angiography performed before EVT (if adequate with sufficiently late venous phases) and considered grades 3 and 4 from the American Society of Interventional and Therapeutic Neuroradiology/Society of Interventional Radiology (ASITN/SIR) classification as good collateral status.27 We also computed the volume of perfusion (Tmax >6 seconds)/volume of the core infarct (diffusion) mismatch to estimate the tissue at risk.

Temporal Evolution of Infarct Volumes

We defined IGR as the ratio of infarct core volume on acute diffusion MRI to the time between stroke onset and imaging. IGR is a simplified metric representing the average growth until imaging that we used as starting point to separate patients with different growth profiles. To do so, we explored the distribution of IGR within the training dataset, and we used an unsupervised clustering approach to identify subtypes of stroke progressors among the entire dataset. We used a Gaussian mixture approach,28 constraining the model to 3 clusters according to the literature,7 later referred to as slow, intermediate, and fast progressors. We first estimated the parameters of the 3 Gaussians (mean values, variances, and mixing coefficients) using the expectation-maximization algorithm under the R package mclust. Then, we hardly assigned each patient to a group based on its highest posterior probability. We explored the different clinical and imaging characteristics between such defined growth groups with Kruskal-Wallis rank sum tests.
Then, we extrapolated pseudo-longitudinal core volume models across time for each group. The rationale behind this approach is that the patients from the same growth phenotype and with the same type of proximal occlusion (all are ICA or M1) should share similar evolution and could be used to extrapolate volumetric trajectories. We have already showed that such concatenation of several cross-sectional data is able to infer the same dynamic as truly longitudinal data.17 We first used the nonparametric locally estimated scatterplot smoothing (LOESS) approach consisting of fitting models to localized subsets of the data point by point29 to represent smooth curves. We also tested 10 models from the simplest to the most complex (linear, logarithmic, inverse, quadratic, cubic, power, S, growth, exponential, and logistic) and kept the one with the highest r value.

Spatial Evolution of Infarct Volumes

We computed 3D frequency maps for 5 consecutive bins of 1 hour in each growth group using the core masks normalized in MNI152 after flipping all the infarcts to the left hemisphere to increase the statistical power. These maps represent the number of time each voxel was affected by an infarct. We quantified the areas affected through frequency map projections into the Harvard-Oxford cortical and subcortical structural atlas.30 We tested the different involvement of the atlas regions according to time, group, and time × group using 2-way multivariate analysis of variance with Wilks' Lambda tests and correction for multiple comparisons.
The previous approach provided the typical infarct growth location according to the growth phenotype but only for 5 steps of delay from stroke onset (of 1 hour each). To provide a smoother stereotypical growth pattern, we also created 3 videos (1 per growth phenotype) from 300 time points uniformly spaced between 0-hour and 5-hour stroke onset. For each time point, we summed all the binary masks (n = 703 in the training dataset), which we weighted by the exponential of the negative distance between the patient volume/time location and its theoretical location based on the best curve fitting for each phenotype described earlier. The resulting weighted volumes were color coded with intensity normalization, stacked cumulatively, and superimposed over the T1 MNI template.

Impact of Growth Group to Anticipate the Clinical Outcome

In the training dataset, we used logistic regression with the growth group as the primary independent variable to predict clinical outcomes, defined as good for a 3-month modified Rankin scale (mRS) ≤2. We checked multicollinearity among variables with Spearman correlations (eFigure 2). We quantified the model performance with the area under the receiver operating characteristic curve (AUC) with bootstrap technique to estimate the 95% CI. Then, we applied this prognosis model to the patients from the independent dataset for external validation. We projected the core volumes of these new patients onto the longitudinal profiles created above to identify their growth groups from the smallest Euclidean distance to 1 of the 3 fitting curves. We compared the probability of these new patients having good outcome with the actual mRS and recomputed AUC and 95% CI. Analyses were implemented on R (version 4.2.2) and SPSS (version 29.0.1.0).

Data Availability

Data could be made available upon request to the principal investigators.

Results

Patient Characteristics

Among the 218 participants enrolled in FRAME, we excluded 93 from this analysis for reasons summarized in the flowchart (Figure 1). Regarding ETIS, from January 2015 to January 2022, we counted 1,747 patients with acute imaging. The images already verified, quality controlled, and uploaded in the storage system (ArchiMed) were available for 1,016 patients. Finally, after application of the inclusion criteria, 578 patients with acute imaging were included and added to the FRAME patients to create the training dataset of 703 patients (Figure 1). During validation analyses (09/2022), an additional set of 101 patients was available from ETIS (Figure 1) and used as an external validation dataset.
The total of 804 patients consisted of 409 (50.9%) men, with a median age of 73.0 years (interquartile range [IQR] 61.2–82.0 years) and a median NIH Stroke Scale (NIHSS) score of 17.0 (IQR 11.0–21.0). Most of the patients (n = 746, 94.1%) were treated with EVT after a median delay of 4.2 hours (IQR 3.1–5.5). Other patient characteristics are summarized in Table 1. The training dataset and the validation dataset showed close characteristics (Table 1).
Table 1 Characteristics of Patients
VariablesAll patients (n = 804)Training dataset (n = 703)Validation dataset (n = 101)
Baseline clinical characteristics   
 Age (y)73.0 (61.2–82)73.0 (61.7–82.0)72.0 (60.0–80.0)
 Sexa   
  Female395 (49.1)344 (48.9)51 (50.5)
  Male409 (50.9)359 (51.1)50 (49.5)
 Hypertensiona461 (58.1)405 (57.9)56 (59.6)
 Diabetes mellitusa122 (15.4)106 (15.2)16 (17.0)
 Hyperlipidemiaa242 (31.2)215 (31.4)27 (29.3)
 Blood glucose level at admission (mM)6.7 (5.7–7.8)6.7 (5.7–7.8)6.8 (5.7–8.2)
 Active smokinga159 (21.2)135 (20.5)24 (26.4)
 History of coronary artery diseasea84 (13.5)69 (13.1)15 (15.8)
 Prior transient ischemic attacka5 (4.0)5 (4.0)NA
 Prior strokea111 (14.2)101 (14.7)10 (10.6)
 Baseline NIHSS score17.0 (11.0–21.0)17.0 (11.0–21.0)17.0 (13.0–21.0)
Baseline radiologic characteristics   
 Time from stroke onset to MRI (h)2.2 (1.6–3.1)2.2 (1.6–3.2)2.1 (1.6–2.7)
 Intracranial ICA occlusiona160 (20.1)140 (20.6)20 (21.5)
 M1 occlusiona620 (77.9)547 (80.7)73 (78.5)
 Both ICA and M1 occlusiona25 (3.1)25 (3.7)0 (0.0)
 Infarct volume (mL)17.5 (6.7–51.2)17.8 (7.2–52.7)15.7 (5.2–39.9)
 Infarct growth rate (mL/h)8.5 (2.7–23.5)8.4 (2.7–24.2)9.3 (2.4–19.6)
 Tissue with Tmax >6 s (mL)99.2 (64.8–133.7)99.2 (64.8–133.7)NA
 Volume of perfusion (Tmax >6 s): diffusion mismatch (mL)61.5 (34.1–91.4)61.5 (34.1–91.4)NA
 Tissue with Tmax >10 s (mL)46.3 (18.4–78.5)46.3 (18.4–78.5)NA
 Hypoperfusion intensity ratio0.4 (0.3–0.5)0.4 (0.3–0.5)NA
 Good collateral statusa210 (43.6)186 (42.7)24 (52.2)
 Hemispherea   
  Left400 (49.8)345 (49.1)55 (54.5)
  Right404 (50.2)358 (50.9)46 (45.5)
Treatments   
 Drip and ship workflowa340 (44.4)313 (46.4)27 (30)
 IV thrombolysisa531 (66.5)464 (66.1)67 (69.1)
 Endovascular thrombectomya746 (94.1)656 (94.3)90 (92.8)
 General anesthesiaa272 (34.7)261 (38.0)11 (11.3)
 Time from stroke onset to end of endovascular thrombectomy (h)4.2 (3.1–5.5)4.2 (3.2–5.5)3.8 (2.9–5.4)
Outcome   
 Successful recanalization, mTICI 2b-3a647 (85.9)562 (85.3)85 (90.4)
 Day 1 hemorrhagic transformation (absent; HI1; HI2; PH1; PH2; PHr; SAH)343 (49.7); 116 (16.8); 96 (13.9); 49 (7.1); 57 (8.3); 6 (0.9); 23 (3.3)302 (48.8); 103 (16.6); 86 (13.9); 46 (7.4); 54 (8.7); 6 (1.0); 22 (3.6)41 (57.8); 13 (18.3); 10 (14.1); 3 (4.2); 3 (4.2); 0 (0.0); 1 (1.4)
 3-month modified Rankin scale3.0 (1.0–4.0)3.0 (1.0–4.0)2.0 (1.0–4.0)
Abbreviations: HI = hemorrhagic infarction; ICA = internal carotid artery; IQR = interquartile range; mTICI = modified treatment in cerebral infarction; NA = not available; NIHSS = NIH Stroke Scale; PH = parenchymal hematoma; PHr = parenchymal hematoma remote; SAH = subarachnoid hemorrhage.
Except where indicated, data are medians, with IQRs in parentheses.
a
Data are numbers of patients, with percentages in parentheses.

Temporal Evolution of Infarct Volumes

Within the training dataset, the median volume of infarct core before recanalization was 17.8 mL (IQR 7.2–52.7), as observed on diffusion MRI collected after a median delay of 2.2 hours (IQR 1.6–3.2, Table 1). This translated into a median calculated IGR of 8.4 mL/h (IQR 2.7–24.2) that showed a non-normal left-skewed distribution (Figure 2A). After applying natural log transformation, we got a more centered distribution, but the IGR spread over a wide range of values, pointing toward different subpopulations within this sample. Under the assumptions of 3 subpopulations, later called slow, intermediate, and fast progressors, we identified a mixture of 3 Gaussians within the overall sample (Figure 2B) that provided the likelihood for each patient to belong to a given group (Figure 2C).
Figure 2 Distribution of Infarct Growth Rate and Temporal Evolution of Infract Volumes
The histogram of infarct growth rate in the training dataset showed a non-normal left-skewed distribution (A). After natural logarithmic transformation, the distribution of infarct growth rate was modeled as a mixture of 3 Gaussian shape subpopulations (B), which allowed the attribution of each patient to a slow, intermediate, or fast progression profile (C). (D) shows the distribution of infarct volumes according to the time from stroke onset for the training dataset. The dynamic evolution of infarct volumes has been modeled through concatenation of the cross-sectional data for the slow, intermediate, and fast progressors. Solid lines show the locally estimated scatterplot smoothing and the associated CIs, while dashed lines show the best parametric fits corresponding to a linear function for the slow progressors (yellow), a cubic function for the intermediate progressors (cyan), and a power function for the fast progressors (purple).
Each patient was then attributed to a group according to its highest probability, which led to separate slow vs intermediate progressors at 0.87 mL/h and intermediate vs fast progressors at 21.47 mL/h and led to a distribution for slow, intermediate, and fast representing 11% (74/703), 62% (437/703), and 27% (192/703), respectively, of the total sample (Table 2).
Table 2 Characteristics of Patients According to the Growth Group in the Training Dataset
VariablesSlow progressors (n = 74, 11%)Intermediate progressors (n = 437, 62%)Fast progressors (n = 192, 27%)p Valuea
Baseline clinical characteristics    
 Age (y)73.0 (66.0–81.5)71 ± 15 (18–98)69 ± 15 (26–98)ns
 Sexb   <0.001
  Female44 (59.5)228 (52.2)72 (37.5) 
  Male30 (40.5)209 (47.8)120 (62.5) 
 Hypertensionb55 (74.3)239 (54.9)111 (58.4)<0.01
 Diabetes mellitusb17 (23.0)56 (12.9)33 (17.4)ns
 Hyperlipidemiab26 (36.1)127 (29.7)62 (33.7)ns
 Blood glucose level at admission (mM)6.7 (5.7–7.5)6.5 (5.7–7.7)6.8 (5.8–8.0)ns
 Active smokingb10 (14.3)77 (18.6)48 (27.3)<0.05
 History of coronary artery diseaseb14 (25.5)34(10.3)21 (14.8)<0.01
 Prior transient ischemic attackb1 (9.1)4 (5.3)0 (0.0)ns
 Prior strokeb13 (18.1)55 (12.9)33 (17.6)ns
 Baseline NIHSS score8.0 (4.0–11.0)16.0 (11.0–20.0)19.0 (16.0–23.0)<0.0001
Baseline radiologic characteristics    
 Time from stroke onset to MRI (h)2.8 (1.7–3.9)2.3 (1.7–3.4)1.8 (1.4–2.3)<0.0001
 Intracranial ICA occlusionb12 (16.2)81 (18.5)47 (24.5)ns
 M1 occlusionb59 (79.7)343 (78.5)145 (75.5)ns
 Both ICA and M1 occlusionb4 (5.4)18 (4.3)3 (1.6)ns
 Infarct volume (mL)0.6 (0.01–1.2)13.1 (7.4–24.6)86.5 (56.6–115.4)<0.0001
 Infarct growth rate (mL/h)0.2 (0.0–0.5)5.8 (2.9–11.5)43.3 (30.5–62.3)<0.0001
 Tissue with Tmax >6 s (mL)67.5 (41.3–87.1)91.3 (63.4–117.4)143.2 (115.9–185.9)<0.0001
 Volume of perfusion (Tmax >6 s): diffusion mismatch (mL)66.6 (38.9–86.6)65.0 (41.0–95.2)51.9 (20.4–77.4)<0.05
 Tissue with Tmax >10 s (mL)9.9 (1.9–28.6)32.8 (14.3–51.0)83.2 (59.4–110.1)<0.0001
 Hypoperfusion intensity ratio0.2 (0.1–0.4)0.4 (0.2–0.5)0.6 (0.5–0.7)<0.0001
 Good collateral statusb34 (73.9)131 (48.9)21 (17.2)<0.0001
 Hemisphereb   ns
  Left40 (54.1)214 (49.0)91 (47.4) 
  Right34 (45.9)223 (51.0)101 (52.6) 
Treatments    
 Drip and ship workflowb32 (45.7)196 (46.7)85 (45.9)ns
 IV thrombolysisb42 (56.8)299 (68.6)123 (64.1)ns
 Endovascular thrombectomyb69 (94.5)406 (93.8)181 (95.3)ns
 General anesthesiab20 (27.4)155 (36.5)74 (39.4)<0.05
 Time from stroke onset to end of endovascular thrombectomy (h)4.7 (3.8–6.5)4.4 (3.2–5.7)3.6 (3.0–4.7)<0.0001
Outcome    
 Successful recanalization, mTICI 2b-3b61 (87.1)357 (87.3)144 (80.0)ns
 Day 1 hemorrhagic transformation (absent; HI1; HI2; PH1; PH2; PHr; SAH)52 (77.6); 11 (16.4); 1 (1.5); 0 (0.0); 1 (1.5); 0 (0.0); 2 (3.0)193 (50.5); 63 (16.5); 47 (12.3); 31 (8.1); 34 (8.9); 5 (1.3); 9 (2.3)57 (33.5); 29 (17.1); 38 (22.4); 15 (8.8); 19 (11.2); 1 (0.6); 11 (6.5)<0.0001
 3-month modified Rankin scale1.0 (0.0–3.0)2.0 (1.0–4.0)4.0 (2.0–6.0)<0.0001
Abbreviations: HI = hemorrhagic infarction; ICA = internal carotid artery; IQR = interquartile range; mTICI = modified treatment in cerebral infarction; NIHSS = NIH Stroke Scale; ns = nonsignificant; PH = parenchymal hematoma; PHr = parenchymal hematoma remote; SAH = subarachnoid hemorrhage.
Except where indicated, data are medians, with IQRs in parentheses.
a
Kruskal-Wallis rank sum test.
b
Data are numbers of patients, with percentages in parentheses.
These growth groups showed significantly different IGR (median 0.2 mL/h [IQR 0.0–0.5 mL/h] vs 5.8 mL/h [IQR 2.9–11.5 mL/h] vs 43.3 mL/h [IQR 30.5–62.3 mL/h]; p < 0.0001) and different initial NIHSS severity (median 8.0 [IQR 4.0–11.0] vs 16.0 [IQR 11.0–20.0] vs 19.0 [IQR 16.0–23.0]; p < 0.0001). A slower growth group was also significantly associated with a higher volume of mismatch and a significantly better collateral status (73.9% of good collaterals vs 48.9% vs 17.2%; p < 0.0001). These patients all showed the same type of anterior LVO and were treated similarly with EVT combined with IV thrombolysis whenever eligible with a nondifferent rate of successful recanalization. However, the rate of high-grade hemorrhagic transformation was significantly higher from slow to intermediate to fast progressors, and the 3-month functional outcome worsened along the groups (median 3-month mRS of 1.0 [IQR 0.0–3.0] vs 2.0 [IQR 1.0–4.0] vs 4.0 [IQR 2.0–6.0]; p < 0.0001).
To highlight the dynamic evolution of stroke, we assumed that these 3 groups can be considered as a homogeneous cluster of patients explored at different time points after the onset of symptoms. Therefore, we inferred the pseudo-longitudinal evolutions of core volumes by fitting the cross-sectional data of each group. LOESS fitting provided the estimated virtual pictures of the expected time course evolution before recanalization according to the growth group (Figure 2D). Because LOESS cannot produce a function that is easily represented by a mathematical formula to apply to new patients' data, we also tested parametric models. The evolution of the slow progressors was best modeled with a low slope linear fitting (r = 0.809; p < 0.001). The evolution of the intermediate progressors was the closest to a cubic function (r = 0.471; p < 0.001), while the logarithmic fit was not far (r = 0.452; p < 0.001). The evolution of the fast progressors was modeled with a power function (r = 0.63; p < 0.001) with values close to an exponential fit (r = 0.558; p < 0.001; Figure 2D, dashed lines).

Spatial Evolution of Infarct Volumes

We generated 3D frequency maps from the coregistered core masks to translate the above-defined time course evolutions into archetype infarct locations according to the growth group. The slow progressors showed no stereotypical infarct pattern but instead small and unpredictable lesions within the ICA territory (Figure 3A). By contrast, the intermediate and fast progressors showed stereotypical patterns (Figure 3, B and C) whose global time courses were significantly different (time × group interaction; p < 0.001; multivariate analysis of covariance). In both groups, the pallidum, putamen, caudate, insular cortex, and frontal operculum cortex were nondifferentially affected from the first hours (time × group interaction; p values close to 1; Figure 4). But the fast progressors were characterized by a quicker and more systematic involvement of the cortical parcels of the ICA territory, including clinically relevant regions such as the precentral and postcentral gyrus and the angular, supramarginal, and the posterior division of the superior temporal gyrus. Figure 4 summarizes the mean proportions of each cortical area affected during time. Video 1 provides a smooth representation of the typical infarct growth that is observed with time passing before recanalization for each growth group at a central representative slice. eFigure 3 shows consistent frequency maps computed from the validation dataset for reproducibility.
Figure 3 Spatial Evolution of Infarct Volumes
Prevalence maps from the training dataset show the frequency of infarct locations for the slow (A), intermediate (B), and fast progressors (C) according to the time from stroke onset divided in 1 hour bin. Color coding indicates the ratio of overlapping lesions of the total number of infarcts in each group and each bin. All the infarct cores have been flipped to the left size for this analysis. The maps are not shown above 5 hours because of the smaller sample size.
Figure 4 Heat Maps of Infarct Location
The 60 cortical and subcortical parcels from the Harvard-Oxford atlas29 are listed on the left. The heat maps show the mean proportions (color coded from 0 to 1) of each parcel that are affected by the infarcts with time passing and according to the growth group in the training dataset. For each parcel, p values indicate the time × group interaction from the 2-way multivariate analysis of variance analysis.
Video 1
Dynamic evolution of infarct volume with time passing according to the growth phenotype. The color indicates infarct location that was computed as the sum of the 703 infarcts (training dataset) weighted by the distance to the fitting curves from the 3 phenotypes and repeatedly calculated for 300 time points uniformly spaced between 0- and 5-hour stroke onset.

Impact of Growth Group to Anticipate the Clinical Outcome

We tested whether the above-defined dynamic evolutions of infarct volume could help anticipating new patients' clinical outcome. Within the training dataset, we found that the growth group was a significant and independent predictor of favorable functional outcome at 3 months (odds ratio [OR] 0.51; 95% CI 0.37–0.72; p < 0.0001) together with age (OR 0.96; 95% CI 0.95–0.97; p < 0.0001) and initial NIHSS severity (OR 0.91; 95% CI 0.88–0.94; p < 0.0001). Overall, we found mean discrimination in terms of AUC of 0.77 (95% CI 0.73–0.80). See Table 3 and eFigure 4A.
Table 3 Logistic Regression Model for the Prediction of Good Functional Outcome (mRS ≤2) on the Training Dataset (n = 703)
VariableUnivariable analysisMultivariable analysis
Odds ratioCorrected p valueOdds ratiop Value
Age (+1 y)0.96 (0.95–0.97)<0.00010.96 (0.95–0.97)<0.0001
Sex (female)0.78 (0.57–1.06)ns  
Hypertension (yes)0.68 (0.49–0.93)ns  
Diabetes mellitus (yes)0.58 (0.37–0.90)ns  
Hyperlipidemia (yes)0.89 (0.64–1.26)ns  
Blood glucose level at admission (+1 mM)0.87 (0.80–0.95)ns  
Active smoking (yes)1.70 (1.13–2.56)ns  
History of coronary artery disease (yes)0.87 (0.50–1.48)ns  
Prior transient ischemic attack (yes)1.43 (0.23–11.12)ns  
Prior stroke (yes)0.51 (0.32–0.81)ns  
Baseline NIHSS score (+1 point)0.88 (0.86–0.91)<0.00010.91 (0.88–0.94)<0.0001
Time from stroke onset to MRI (h)1.07 (0.10–1.15)ns  
Intracranial ICA occlusion (yes)0.80 (0.58–1.09)ns  
M1 occlusion (yes)1.49 (1.04–2.15)ns  
Infarct volume (+1 mL)0.98 (0.98–0.99)<0.0001  
Infarct growth rate (+1 mL/h)0.97 (0.97–0.98)<0.0001  
Growth group (slow, intermediate, and fast)0.41 (0.31–0.54)<0.00010.51 (0.37–0.72)<0.0001
Tissue with Tmax >6 s (+1 mL)0.99 (0.99–1.00)<0.05  
Volume of perfusion (Tmax >6 s): diffusion mismatch (+1 mL)1.00 (0.99–1.00)ns  
Tissue with Tmax >10 s (+1 mL)0.99 (0.98–1.00)<0.05  
Hypoperfusion intensity ratio0.22 (0.06–0.75)ns  
Good collateral status (yes)1.74 (1.17–2.60)ns  
Drip and ship workflow (yes)0.76 (0.55–1.04)ns  
IV thrombolysis (yes)1.93 (1.38–2.70)<0.01  
Endovascular thrombectomy (yes)0.92 (0.45–1.89)ns  
General anesthesia (yes)0.10 (0.74–1.34)ns  
Time from stroke onset to end of endovascular thrombectomy (+1 h)0.99 (0.93–1.05)ns  
Successful recanalization, mTICI 2b-3 (yes)3.05 (1.87–5.16)<0.001  
Day 1 hemorrhagic transformation (no)3.66 (2.60–5.20)<0.0001  
Abbreviations: ICA = internal carotid artery; mRS = modified Rankin scale; mTICI = modified treatment in cerebral infarction; NIHSS = NIH Stroke Scale; ns = nonsignificant.
Data in parentheses are 95% CIs. In univariable analysis, p values have been corrected for multiple comparisons (Bonferroni corrected for 29 comparisons). In the multivariable analysis, we kept only the variables that could be available at the patient admission, which is not the case for the variables related to the type of treatment or its outcome (IV thrombolysis, endovascular thrombectomy, general anesthesia, time from onset to the end of endovascular thrombectomy, successful recanalization, and hemorrhagic transformation). There was a significant collinearity between the variables growth group, infarct volume, infarct growth, and tissue with Tmax >6 seconds, which was expected. We favored the growth group in the multivariable analysis because it is the primary interest of this analysis.
The validation dataset included 87 patients who had an mRS at 3 months from the 101 available. Within this sample that was not used to build the above-mentioned predictive model, we could project each patient onto the stereotypical time course evolutions defined in Figure 2D, which allowed us to attribute a growth group to each of these new patients according to the smallest Euclidean distance with 1 of the 3 infarct growth fits. We classified 34% (30/87) of patients as slow progressors, 51% (44/87) as intermediate, and 15% (13/87) as fast progressors. Using these groups and the previously defined predictive model, we predicted mRS at 3 months with an AUC of 0.78 (95% CI 0.66–0.88). The calibration plot showed fairly good correspondence between the predicted probability of good outcome and the observed outcome (eFigure 4, B and C).

Discussion

The individual tolerance to ischemia following LVO varies significantly among patients, leading to the recent terminology of slow or fast progressors. But no consensual definitions exist for such phenotypes whose dynamic courses needed to be discovered. In this study, we used a data-driven nonsupervised approach to delineate distinct categories of stroke progressors: slow, intermediate, and fast, and we unveiled the characteristic dynamic evolutions of these groups in both time (with volume along time charts) and space (with 3D frequency maps), while shedding light on their clinical relevance. Indeed, using these charts, new patients from an independent validation dataset could be classified in a progressor category, and their probabilities of good outcome could be predicted. In the future, this may help decide therapeutic strategies and may guide patient selection criteria for future neuroprotective trials.
The ambiguity surrounding the definition of slow and fast progressors has been a notable challenge to transfer this otherwise simple concept into clinical practice. For instance, fast has been defined as a core volume >70 mL within the first 6 hours and slow <30 mL between 6 and 24 hours,7,13 but also according to the evolution of the Alberta Stroke Programme Early CT score,9,10,31 to dichotomy of IGR based on the median value within a sample,12,32 to tertile,11 or to a value that best correlates with the clinical outcome.8 To address this issue, we opted for a data-driven, nonsupervised approach using a soft clustering mixture model. This methodology avoids reliance on arbitrary thresholds and instead identifies groups that bear similarities to those described in previous literature. Notably, our results align with existing estimates, suggesting that approximately 20%–30% of patients with LVOS may fall into a fast-to-ultra-fast progressor category,6 a proportion consistent with our 27% fast progressors. Our intermediate group could resemble those who were previously categorized as slow progressors, while we identified a smaller subset (∼11%) as slow progressors who exhibit minimal growth despite proximal occlusion.
One significant contribution of our study lies in using more than 700 core volumes, all coregistered within the same standard space, to infer longitudinal profiles of untreated stroke progression, drawing inspiration from methodologies applied in other domains.15-18 In contrast to previous efforts that used linear fitting,33 which simplifies the complex biological processes at play, we used smooth curve fitting. This approach revealed the nonlinear nature of infarct growth. Our observations were consistent with serial MRI scans in animal stroke models, which indicated a logarithmic pattern for the natural evolution of infarct volumes.34,35 In the case of fast progressors, results indicated that a power function provided a better fit, reflecting a steeper initial increase than the logarithmic pattern. In addition, we noted that an exponential shape could describe the trajectory of fast progressors in the initial hours but failed to account for asymptomatic growth linked to the maximum volume of the vascular territory. However, we acknowledge that no model could describe the data perfectly, which we attribute to some remaining interindividual variability inherent to the concatenation of cross-sectional data, even after clustering. Further investigations, such as trajectory-based clustering36 might be beneficial when leveraging more extensive databases from international networks and data-sharing initiatives.
Our 3D frequency maps and the dynamic video stack provided a spatial dimension to the temporal progression of stroke, offering insights into regions that exhibit varying vulnerability or tolerance to ischemia. We could revisit the known susceptibility of the deep nuclei and the insular cortex, which is in line with frail pial collateral support of these regions.37 Indeed, collateral status emerged as a key determinant of our growth profiles, a finding consistent with previous research.3,4
Furthermore, our growth group classification proved valuable in predicting functional outcomes. The predicted probability of good outcome of patients from the validation dataset were close to the observed one, with an AUC of 0.78 (95% CI 0.66–0.88). It is, however, essential to acknowledge the inherent challenges in making acute-stage predictions with only pretreatment information. However, we believe the method to be sufficiently accurate to capture groups with homogeneous expected outcome, which could help to maximize statistical power in future cerebroprotectant trials. Indeed, an effective strategy interfering with infarct growth before EVT (such as during transfer to a comprehensive stroke center) will have a higher expected effect on fast progressors than on the other phenotypes, which should translate into statistical demonstration of its efficacy with a smaller number of participants if only fast progressors are included vs unselected patients.
Several limitations warrant consideration. We focused on patients considered eligible for EVT, potentially excluding certain patients, particularly in the fast progressor group. Nonetheless, our inclusion of close to 30% of fast progressors surpassed the proportion within initial clinical trials that often involve favorable profiles.38 In that sense, the ETIS real-life prospective registry, representing the largest reported population from multicenter acquisitions, is a strength for translating our results to daily clinical practice. The utilization of MRI limited the generalizability of our findings to settings where CT is the primary imaging modality. However, this approach, which is the first-line imaging in several French centers, together with the inclusion of witnessed symptom onset only, which was not systematic in prior works,8,11 provided high accuracy for growth quantifications. We used the Gaussian mixture as a first approach to provide more objective attribution to a growth group, but several alternative clustering approaches exist and could be tested in the future. We also acknowledge the need for collective efforts in the future to incorporate more data from extended time windows to generalize our IGR charts because most of our patients presented within 3 hours from the onset. This will also be important to demonstrate that the combination of infarct volume and time from onset into a growth rate metric is more relevant than the using each separately. We will also have to explore the impact of recanalization on spatiotemporal dynamics and identify the growth profiles associated with different clot locations. In addition, we based our external validation on the clinical outcome because we could not access a natural history follow-up imaging to check whether the infarct growth followed the prediction. This could be tested in the future with subgroup of patients whose recanalization failed. Finally, PWI was not always available, and therefore, we could not compute HIR for all the patients, and we also assessed collateral status through conventional angiography when PWI was not available. We acknowledge that pre-thrombectomy angiographies don't necessarily include late series and that combining HIR and ASITN/SIR is suboptimal, although several papers showed that both metrics are correlated.24-26 But although our collaterality metric is imperfect, the association of collateral status with the growth phenotypes was already reported4 and was not the main objective of our work that rather focused on revealing unbiased growth charts for infract progression in time and space.
In conclusion, our study has unveiled infarct growth charts for stroke progression that offer a novel and practical approach for phenotype identification in stroke patients and outcome estimation, all without relying on advanced postprocessing techniques. In the future, these growth charts could aid the design of neuroprotective trials and to personalize the treatment strategy.

Glossary

ADC
apparent diffusion coefficient
ASITN/SIR
American Society of Interventional and Therapeutic Neuroradiology/Society of Interventional Radiology
AUC
area under the receiver operating characteristic curve
EVT
endovascular thrombectomy
HIR
hypoperfusion intensity ratio
ICA
internal carotid artery
IGR
infarct growth rate
IQR
interquartile range
LOESS
locally estimated scatterplot smoothing
LVO
large vessel occlusion
LVOS
large vessel occlusion stroke
MNI152
Montreal Neurologic Institute 152
mRS
modified Rankin scale
NIHSS
NIH Stroke Scale
OR
odds ratio
PWI
perfusion-weighted imaging
Tmax
time-to-maximum

Appendix 1 Authors

NameLocationContribution
Fanny Munsch, PhDInstitut de Bio-imagerie IBIO, Univ. Bordeaux, FranceDrafting/revision of the article for content, including medical writing for content; analysis or interpretation of data
David Planes, MDNeuroimagerie Diagnostique et Thérapeutique, CHU de Bordeaux, FranceDrafting/revision of the article for content, including medical writing for content; analysis or interpretation of data
Hikaru Fukutomi, MD, PhDKansai Electric Power Hospital, Osaka, JapanDrafting/revision of the article for content, including medical writing for content; analysis or interpretation of data
Gaultier Marnat, MDNeuroimagerie Diagnostique et Thérapeutique, CHU de Bordeaux, FranceDrafting/revision of the article for content, including medical writing for content; major role in the acquisition of data
Thomas Courret, MDNeuroimagerie Diagnostique et Thérapeutique, CHU de Bordeaux, FranceDrafting/revision of the article for content, including medical writing for content; major role in the acquisition of data
Emilien Micard, PhDInserm CIC-IT U1433, CHRU Nancy, FranceDrafting/revision of the article for content, including medical writing for content; major role in the acquisition of data
Bailiang Chen, PhDInserm CIC-IT U1433, CHRU Nancy, FranceDrafting/revision of the article for content, including medical writing for content; major role in the acquisition of data
Pierre Seners, MD, PhDInstitut de Psychiatrie et Neurosciences de Paris (IPNP), INSERM U1266; Département de Neurologie, Hopital Fondation Rothschild, Paris, FranceDrafting/revision of the article for content, including medical writing for content; study concept or design
Johanna Dubos, MScInstitut de Bio-imagerie IBIO, Univ. Bordeaux, FranceDrafting/revision of the article for content, including medical writing for content; analysis or interpretation of data
Vincent Planche, MD, PhDInstitut des Maladies Neurodégénératives, CNRS, UMR 5293, Univ. Bordeaux, FranceDrafting/revision of the article for content, including medical writing for content; study concept or design
Pierrick Coupé, PhDBordeaux INP, LABRI, CNRS, UMR5800, Univ. Bordeaux, FranceDrafting/revision of the article for content, including medical writing for content; study concept or design
Vincent Dousset, MD, PhDInstitut de Bio-imagerie IBIO, Univ. Bordeaux; Neuroimagerie Diagnostique et Thérapeutique, CHU de Bordeaux; Neurocentre Magendie, INSERM U1215, Univ. Bordeaux, FranceDrafting/revision of the article for content, including medical writing for content
Bertrand Lapergue, MD, PhDService de Neurologie et Unité de Neuro Vasculaire, Hôpital FOCH, Suresnes, FranceDrafting/revision of the article for content, including medical writing for content; major role in the acquisition of data
Jean Marc Olivot, MD, PhDUnité Neurovasculaire, CHU de Toulouse, FranceDrafting/revision of the article for content, including medical writing for content; major role in the acquisition of data
Igor Sibon, MD, PhDUnité Neurovasculaire, CHU de Bordeaux, FranceDrafting/revision of the article for content, including medical writing for content; major role in the acquisition of data; and study concept or design
Michel Thiebaut De Schotten, PhDCNRS, UMR-5293, Univ. Bordeaux; Brain Connectivity and Behaviour Laboratory, Paris, FranceDrafting/revision of the article for content, including medical writing for content; study concept or design; and analysis or interpretation of data
Thomas Tourdias, MD, PhDInstitut de Bio-imagerie IBIO, Univ. Bordeaux; Neuroimagerie Diagnostique et Thérapeutique, CHU de Bordeaux; Neurocentre Magendie, INSERM U1215, Univ. Bordeaux, FranceDrafting/revision of the article for content, including medical writing for content; major role in the acquisition of data; study concept or design; and analysis or interpretation of data

Appendix 2 Coinvestigators

Coinvestigators are listed at Neurology.org.

Supplementary Material

File (supplementary_coinvestigator.pdf)
File (supplementary_figure.pdf)
File (supplementary_table1.pdf)

References

1.
Fisher M, Savitz SI. Pharmacological brain cytoprotection in acute ischaemic stroke: renewed hope in the reperfusion era. Nat Rev Neurol. 2022;18(4):193-202.
Crossref
PubMed
Google Scholar
2.
Desai SM, Rocha M, Jovin TG, Jadhav AP. High variability in neuronal loss. Stroke. 2019;50(1):34-37.
Crossref
PubMed
Google Scholar
3.
Vagal A, Aviv R, Sucharew H, et al. Collateral clock is more important than time clock for tissue fate. Stroke. 2018;49(9):2102-2107.
Crossref
PubMed
Google Scholar
4.
Seners P, Yuen N, Olivot JM, et al. Factors associated with fast early infarct growth in patients with acute ischemic stroke with a large vessel occlusion. Neurology. 2023;101(21):e2126-e2137.
Crossref
PubMed
Google Scholar
5.
Powers WJ, Rabinstein AA, Ackerson T, et al. 2018 guidelines for the early management of patients with acute ischemic stroke: a guideline for healthcare professionals from the American Heart Association/American Stroke Association. Stroke. 2018;49(3):e46-e110.
Crossref
PubMed
Google Scholar
6.
Rocha M, Jovin TG. Fast versus slow progressors of infarct growth in large vessel occlusion stroke: clinical and research implications. Stroke. 2017;48(9):2621-2627.
Crossref
PubMed
Google Scholar
7.
Rocha M, Desai SM, Jadhav AP, Jovin TG. Prevalence and temporal distribution of fast and slow progressors of infarct growth in large vessel occlusion stroke. Stroke. 2019;50(8):2238-2240.
Crossref
PubMed
Google Scholar
8.
Sarraj A, Hassan AE, Grotta J, et al. Early infarct growth rate correlation with endovascular thrombectomy clinical outcomes: analysis from the SELECT study. Stroke. 2021;52(1):57-69.
Crossref
PubMed
Google Scholar
9.
Marnat G, Kaesmacher J, Buetikofer L, et al. Interaction between intravenous thrombolysis and clinical outcome between slow and fast progressors undergoing mechanical thrombectomy: a post-hoc analysis of the SWIFT-DIRECT trial. J Neurointerv Surg. 2023;16(1):45-52.
Crossref
PubMed
Google Scholar
10.
Ospel JM, Hill MD, Kappelhof M, et al. Which acute ischemic stroke patients are fast progressors? Results from the ESCAPE trial control arm. Stroke. 2021;52(5):1847-1850.
Crossref
PubMed
Google Scholar
11.
Seo WK, Liebeskind DS, Yoo B, et al. Predictors and functional outcomes of fast, intermediate, and slow progression among patients with acute ischemic stroke. Stroke. 2020;51(8):2553-2557.
Crossref
PubMed
Google Scholar
12.
Mohammaden MH, Haussen DC, Pisani L, et al. Characterizing fast and slow progressors in anterior circulation large vessel occlusion strokes. Interv Neuroradiol. 2023;29(4):379-385.
Crossref
PubMed
Google Scholar
13.
Rocha M, Desai S, Son J, Tonetti DA, Jovin T, Jadhav AP. Clinical characteristics of fast and slow progressors of infarct growth in anterior circulation large vessel occlusion stroke. J Cereb Blood Flow Metab. 2021;41(7):271678X211015068.
Crossref
PubMed
Google Scholar
14.
Seners P, Scheldeman L, Christensen S, et al. Determinants of infarct core growth during inter-hospital transfer for thrombectomy. Ann Neurol. 2023;93(6):1117-1129.
Crossref
PubMed
Google Scholar
15.
Bethlehem RAI, Seidlitz J, White SR, et al. Brain charts for the human lifespan. Nature. 2022;604(7906):525-533.
Crossref
PubMed
Google Scholar
16.
Coupe P, Catheline G, Lanuza E, Manjon JV; Alzheimer's Disease Neuroimaging Initiative. Towards a unified analysis of brain maturation and aging across the entire lifespan: a MRI analysis. Hum Brain Mapp. 2017;38(11):5501-5518.
Crossref
PubMed
Google Scholar
17.
Planche V, Manjon JV, Mansencal B, et al. Structural progression of Alzheimer's disease over decades: the MRI staging scheme. Brain Commun. 2022;4(3):fcac109.
Crossref
PubMed
Google Scholar
18.
Planche V, Mansencal B, Manjon JV, et al. Anatomical MRI staging of frontotemporal dementia variants. Alzheimers Dement. 2023;19(8):3283-3294.
Crossref
PubMed
Google Scholar
19.
von Elm E, Altman DG, Egger M, et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Int J Surg. 2014;12(12):1495-1499.
Crossref
PubMed
Google Scholar
20.
Olivot JM, Albucher JF, Guenego A, et al. Mismatch profile influences outcome after mechanical thrombectomy. Stroke. 2021;52(1):232-240.
Crossref
PubMed
Google Scholar
21.
Fukutomi H, Yamamoto T, Sibon I, et al. Location-weighted versus volume-weighted mismatch at MRI for response to mechanical thrombectomy in acute stroke. Radiology. 2023;306(2):e220080.
Crossref
PubMed
Google Scholar
22.
Maier B, Finitsis S, Mazighi M, et al. The benefit of a complete over a successful reperfusion decreases with time. Ann Neurol. 2023;93(5):934-941.
Crossref
PubMed
Google Scholar
23.
Grech-Sollars M, Hales PW, Miyazaki K, et al. Multi-centre reproducibility of diffusion MRI parameters for clinical sequences in the brain. NMR Biomed. 2015;28(4):468-485.
Crossref
PubMed
Google Scholar
24.
Olivot JM, Mlynash M, Inoue M, et al. Hypoperfusion intensity ratio predicts infarct progression and functional outcome in the DEFUSE 2 Cohort. Stroke. 2014;45(4):1018-1023.
Crossref
PubMed
Google Scholar
25.
Guenego A, Fahed R, Albers GW, et al. Hypoperfusion intensity ratio correlates with angiographic collaterals in acute ischaemic stroke with M1 occlusion. Eur J Neurol. 2020;27(5):864-870.
Crossref
PubMed
Google Scholar
26.
Ai Z, Jiang L, Zhao B, Su H, Yin X, Chen YC. Hypoperfusion intensity ratio correlates with angiographic collaterals and infarct growth in acute stroke with thrombectomy. Curr Med Imaging. 2023;19(13):1561-1569.
Crossref
PubMed
Google Scholar
27.
Zaidat OO, Yoo AJ, Khatri P, et al. Recommendations on angiographic revascularization grading standards for acute ischemic stroke: a consensus statement. Stroke. 2013;44(9):2650-2663.
Crossref
PubMed
Google Scholar
28.
Figueiredo M, Jain A. Unsupervised learning of finite mixture models. IEEE Transact Pattern Anal Mach Intell. 2002;24(3):381-396.
Crossref
Google Scholar
29.
Cleveland S, Devlin S. Locally weighted regression: an approach to regression analysis by local fitting. J Am Stat Assoc. 1988;83(403):596-610.
Crossref
Google Scholar
30.
Desikan RS, Segonne F, Fischl B, et al. An automated labeling system for subdividing the human cerebral cortex on MRI scans into gyral based regions of interest. Neuroimage. 2006;31(3):968-980.
Crossref
PubMed
Google Scholar
31.
Reddy ST, Friedman E, Wu TC, et al. Rapid infarct progression in anterior circulation large vessel occlusion ischemic stroke patients during inter-facility transfer. J Stroke Cerebrovasc Dis. 2020;29(12):105308.
Crossref
PubMed
Google Scholar
32.
Renú Jornet A, Urra X, Laredo C, et al. Benefit from mechanical thrombectomy in acute ischemic stroke with fast and slow progression. J Neurointerv Surg. 2020;12(2):132-135.
Crossref
PubMed
Google Scholar
33.
Seners P, Yuen N, Mlynash M, et al. Quantification of penumbral volume in association with time from stroke onset in acute ischemic stroke with large vessel occlusion. JAMA Neurol. 2023;80(5):523-528.
Crossref
PubMed
Google Scholar
34.
Zhang X, Tong F, Li CX, et al. Temporal evolution of ischemic lesions in nonhuman primates: a diffusion and perfusion MRI study. PLoS One. 2015;10(2):e0117290.
Crossref
PubMed
Google Scholar
35.
Christoforidis GA, Vakil P, Ansari SA, Dehkordi FH, Carroll TJ. Impact of pial collaterals on infarct growth rate in experimental acute ischemic stroke. AJNR Am J Neuroradiol. 2017;38(2):270-275.
Crossref
PubMed
Google Scholar
36.
Den Teuling NGP, Pauws SC, van den Heuvel ER. A comparison of methods for clustering longitudinal data with slowly changing trends. Commun Stat Simul Comput. 2023;52(3):621-648.
Crossref
Google Scholar
37.
Vogels V, Dammers R, van Bilsen M, Volovici V. Deep cerebral perforators: anatomical distribution and clinical symptoms: an overview. Stroke. 2021;52(10):e660-e674.
Crossref
PubMed
Google Scholar
38.
Albers GW, Marks MP, Kemp S, et al. Thrombectomy for stroke at 6 to 16 hours with selection by perfusion imaging. N Engl J Med. 2018;378(8):708-718.
Crossref
PubMed
Google Scholar

Information & Authors

Information

Published In

Neurology®
Volume 102 • Number 12 • June 25, 2024
PubMed: 38815232

Copyright

© 2024 American Academy of Neurology.

Publication History

Received: December 14, 2023
Accepted: March 4, 2024
Published online: May 30, 2024
Published in issue: June 25, 2024

Permissions

Request permissions for this article.

Disclosure

The authors report no relevant disclosures. Go to Neurology.org/N for full disclosures.

Study Funding

This work received financial support from the French government in the framework of the University of Bordeaux's France 2030 program/RRI IMPACT and IHU Precision & Global Vascular Brain Health Institute - VBHI. FRAME was funded by a public grant from the French Ministry of Health, Clinical Research Hospital Program 2015 (PHRCI-15-076). This work was supported by the European Union's Horizon 2020 research and innovation program under the European Research Council (ERC) Consolidator grant agreement no. 818521 (M.T.d.S., DISCONNECTOME).

Authors

Affiliations & Disclosures

Fanny Munsch, PhD* https://orcid.org/0000-0002-7144-8617
From the Institut de Bio-imagerie IBIO (F.M., J.D., V.D., T.T.), University Bordeaux; Neuroimagerie Diagnostique et Thérapeutique (D.P., G.M., T.C., V.D., T.T.), CHU de Bordeaux, France; Kansai Electric Power Hospital (H.F.), Osaka, Japan; Inserm CIC-IT U1433 (E.M., B.C.), CHRU Nancy; Institut de Psychiatrie et Neurosciences de Paris (IPNP) (P.S.), INSERM U1266; Département de Neurologie (P.S.), Hopital Fondation Rothschild, Paris; Institut des Maladies Neurodégénératives (V.P.), CNRS, UMR 5293, Bordeaux INP (P.C.), LABRI, CNRS, UMR5800, and Neurocentre Magendie (V.D., T.T.), INSERM U1215, Univ. Bordeaux; Service de Neurologie et Unité de Neuro Vasculaire (B.L.), Hôpital FOCH, Suresnes; Unité Neurovasculaire (J.M.O.), CHU de Toulouse; Unité Neurovasculaire (I.S.), CHU de Bordeaux; CNRS (M.T.D.S.), UMR-5293, Univ. Bordeaux; and Brain Connectivity and Behaviour Laboratory (M.T.D.S.), Paris, France.
Disclosure
Financial Disclosure:
1.
NONE
Research Support:
1.
NONE
Stock, Stock Options & Royalties:
1.
NONE
Legal Proceedings:
1.
NONE
View all articles by this author
David Planes, MD*
From the Institut de Bio-imagerie IBIO (F.M., J.D., V.D., T.T.), University Bordeaux; Neuroimagerie Diagnostique et Thérapeutique (D.P., G.M., T.C., V.D., T.T.), CHU de Bordeaux, France; Kansai Electric Power Hospital (H.F.), Osaka, Japan; Inserm CIC-IT U1433 (E.M., B.C.), CHRU Nancy; Institut de Psychiatrie et Neurosciences de Paris (IPNP) (P.S.), INSERM U1266; Département de Neurologie (P.S.), Hopital Fondation Rothschild, Paris; Institut des Maladies Neurodégénératives (V.P.), CNRS, UMR 5293, Bordeaux INP (P.C.), LABRI, CNRS, UMR5800, and Neurocentre Magendie (V.D., T.T.), INSERM U1215, Univ. Bordeaux; Service de Neurologie et Unité de Neuro Vasculaire (B.L.), Hôpital FOCH, Suresnes; Unité Neurovasculaire (J.M.O.), CHU de Toulouse; Unité Neurovasculaire (I.S.), CHU de Bordeaux; CNRS (M.T.D.S.), UMR-5293, Univ. Bordeaux; and Brain Connectivity and Behaviour Laboratory (M.T.D.S.), Paris, France.
Disclosure
Financial Disclosure:
1.
NONE
Research Support:
1.
NONE
Stock, Stock Options & Royalties:
1.
NONE
Legal Proceedings:
1.
NONE
View all articles by this author
Hikaru Fukutomi, MD, PhD
From the Institut de Bio-imagerie IBIO (F.M., J.D., V.D., T.T.), University Bordeaux; Neuroimagerie Diagnostique et Thérapeutique (D.P., G.M., T.C., V.D., T.T.), CHU de Bordeaux, France; Kansai Electric Power Hospital (H.F.), Osaka, Japan; Inserm CIC-IT U1433 (E.M., B.C.), CHRU Nancy; Institut de Psychiatrie et Neurosciences de Paris (IPNP) (P.S.), INSERM U1266; Département de Neurologie (P.S.), Hopital Fondation Rothschild, Paris; Institut des Maladies Neurodégénératives (V.P.), CNRS, UMR 5293, Bordeaux INP (P.C.), LABRI, CNRS, UMR5800, and Neurocentre Magendie (V.D., T.T.), INSERM U1215, Univ. Bordeaux; Service de Neurologie et Unité de Neuro Vasculaire (B.L.), Hôpital FOCH, Suresnes; Unité Neurovasculaire (J.M.O.), CHU de Toulouse; Unité Neurovasculaire (I.S.), CHU de Bordeaux; CNRS (M.T.D.S.), UMR-5293, Univ. Bordeaux; and Brain Connectivity and Behaviour Laboratory (M.T.D.S.), Paris, France.
Disclosure
Financial Disclosure:
1.
NONE
Research Support:
1.
NONE
Stock, Stock Options & Royalties:
1.
NONE
Legal Proceedings:
1.
NONE
View all articles by this author
Gaultier Marnat, MD https://orcid.org/0000-0002-7611-7753
From the Institut de Bio-imagerie IBIO (F.M., J.D., V.D., T.T.), University Bordeaux; Neuroimagerie Diagnostique et Thérapeutique (D.P., G.M., T.C., V.D., T.T.), CHU de Bordeaux, France; Kansai Electric Power Hospital (H.F.), Osaka, Japan; Inserm CIC-IT U1433 (E.M., B.C.), CHRU Nancy; Institut de Psychiatrie et Neurosciences de Paris (IPNP) (P.S.), INSERM U1266; Département de Neurologie (P.S.), Hopital Fondation Rothschild, Paris; Institut des Maladies Neurodégénératives (V.P.), CNRS, UMR 5293, Bordeaux INP (P.C.), LABRI, CNRS, UMR5800, and Neurocentre Magendie (V.D., T.T.), INSERM U1215, Univ. Bordeaux; Service de Neurologie et Unité de Neuro Vasculaire (B.L.), Hôpital FOCH, Suresnes; Unité Neurovasculaire (J.M.O.), CHU de Toulouse; Unité Neurovasculaire (I.S.), CHU de Bordeaux; CNRS (M.T.D.S.), UMR-5293, Univ. Bordeaux; and Brain Connectivity and Behaviour Laboratory (M.T.D.S.), Paris, France.
Disclosure
Financial Disclosure:
1.
Personal Compensation: (1) Consultant - Stryker Neurovascular (2) Consultant - Microvention Europe (3) Consultant - Balt (4) Paid lecture - Phenox (5) Paid lecture - Cerenovus (6) Paid lecture - Bracco (7) Paid lecture - Medtronic (8) Paid lecture - Penumbra
Research Support:
1.
NONE
Stock, Stock Options & Royalties:
1.
NONE
Legal Proceedings:
1.
NONE
View all articles by this author
Thomas Courret, MD https://orcid.org/0000-0002-7282-1153
From the Institut de Bio-imagerie IBIO (F.M., J.D., V.D., T.T.), University Bordeaux; Neuroimagerie Diagnostique et Thérapeutique (D.P., G.M., T.C., V.D., T.T.), CHU de Bordeaux, France; Kansai Electric Power Hospital (H.F.), Osaka, Japan; Inserm CIC-IT U1433 (E.M., B.C.), CHRU Nancy; Institut de Psychiatrie et Neurosciences de Paris (IPNP) (P.S.), INSERM U1266; Département de Neurologie (P.S.), Hopital Fondation Rothschild, Paris; Institut des Maladies Neurodégénératives (V.P.), CNRS, UMR 5293, Bordeaux INP (P.C.), LABRI, CNRS, UMR5800, and Neurocentre Magendie (V.D., T.T.), INSERM U1215, Univ. Bordeaux; Service de Neurologie et Unité de Neuro Vasculaire (B.L.), Hôpital FOCH, Suresnes; Unité Neurovasculaire (J.M.O.), CHU de Toulouse; Unité Neurovasculaire (I.S.), CHU de Bordeaux; CNRS (M.T.D.S.), UMR-5293, Univ. Bordeaux; and Brain Connectivity and Behaviour Laboratory (M.T.D.S.), Paris, France.
Disclosure
Financial Disclosure:
1.
NONE
Research Support:
1.
NONE
Stock, Stock Options & Royalties:
1.
NONE
Legal Proceedings:
1.
NONE
View all articles by this author
Emilien Micard, PhD
From the Institut de Bio-imagerie IBIO (F.M., J.D., V.D., T.T.), University Bordeaux; Neuroimagerie Diagnostique et Thérapeutique (D.P., G.M., T.C., V.D., T.T.), CHU de Bordeaux, France; Kansai Electric Power Hospital (H.F.), Osaka, Japan; Inserm CIC-IT U1433 (E.M., B.C.), CHRU Nancy; Institut de Psychiatrie et Neurosciences de Paris (IPNP) (P.S.), INSERM U1266; Département de Neurologie (P.S.), Hopital Fondation Rothschild, Paris; Institut des Maladies Neurodégénératives (V.P.), CNRS, UMR 5293, Bordeaux INP (P.C.), LABRI, CNRS, UMR5800, and Neurocentre Magendie (V.D., T.T.), INSERM U1215, Univ. Bordeaux; Service de Neurologie et Unité de Neuro Vasculaire (B.L.), Hôpital FOCH, Suresnes; Unité Neurovasculaire (J.M.O.), CHU de Toulouse; Unité Neurovasculaire (I.S.), CHU de Bordeaux; CNRS (M.T.D.S.), UMR-5293, Univ. Bordeaux; and Brain Connectivity and Behaviour Laboratory (M.T.D.S.), Paris, France.
Disclosure
Financial Disclosure:
1.
NONE
Research Support:
1.
NONE
Stock, Stock Options & Royalties:
1.
NONE
Legal Proceedings:
1.
NONE
View all articles by this author
Bailiang Chen, PhD https://orcid.org/0000-0003-1038-7846
From the Institut de Bio-imagerie IBIO (F.M., J.D., V.D., T.T.), University Bordeaux; Neuroimagerie Diagnostique et Thérapeutique (D.P., G.M., T.C., V.D., T.T.), CHU de Bordeaux, France; Kansai Electric Power Hospital (H.F.), Osaka, Japan; Inserm CIC-IT U1433 (E.M., B.C.), CHRU Nancy; Institut de Psychiatrie et Neurosciences de Paris (IPNP) (P.S.), INSERM U1266; Département de Neurologie (P.S.), Hopital Fondation Rothschild, Paris; Institut des Maladies Neurodégénératives (V.P.), CNRS, UMR 5293, Bordeaux INP (P.C.), LABRI, CNRS, UMR5800, and Neurocentre Magendie (V.D., T.T.), INSERM U1215, Univ. Bordeaux; Service de Neurologie et Unité de Neuro Vasculaire (B.L.), Hôpital FOCH, Suresnes; Unité Neurovasculaire (J.M.O.), CHU de Toulouse; Unité Neurovasculaire (I.S.), CHU de Bordeaux; CNRS (M.T.D.S.), UMR-5293, Univ. Bordeaux; and Brain Connectivity and Behaviour Laboratory (M.T.D.S.), Paris, France.
Disclosure
Financial Disclosure:
1.
NONE
Research Support:
1.
NONE
Stock, Stock Options & Royalties:
1.
NONE
Legal Proceedings:
1.
NONE
View all articles by this author
Pierre Seners, MD, PhD https://orcid.org/0000-0002-2134-0691
From the Institut de Bio-imagerie IBIO (F.M., J.D., V.D., T.T.), University Bordeaux; Neuroimagerie Diagnostique et Thérapeutique (D.P., G.M., T.C., V.D., T.T.), CHU de Bordeaux, France; Kansai Electric Power Hospital (H.F.), Osaka, Japan; Inserm CIC-IT U1433 (E.M., B.C.), CHRU Nancy; Institut de Psychiatrie et Neurosciences de Paris (IPNP) (P.S.), INSERM U1266; Département de Neurologie (P.S.), Hopital Fondation Rothschild, Paris; Institut des Maladies Neurodégénératives (V.P.), CNRS, UMR 5293, Bordeaux INP (P.C.), LABRI, CNRS, UMR5800, and Neurocentre Magendie (V.D., T.T.), INSERM U1215, Univ. Bordeaux; Service de Neurologie et Unité de Neuro Vasculaire (B.L.), Hôpital FOCH, Suresnes; Unité Neurovasculaire (J.M.O.), CHU de Toulouse; Unité Neurovasculaire (I.S.), CHU de Bordeaux; CNRS (M.T.D.S.), UMR-5293, Univ. Bordeaux; and Brain Connectivity and Behaviour Laboratory (M.T.D.S.), Paris, France.
Disclosure
Financial Disclosure:
1.
Personal Compensation: (1) Speaker honoraria - Acticor biotech
Research Support:
1.
NONE
Stock, Stock Options & Royalties:
1.
NONE
Legal Proceedings:
1.
NONE
View all articles by this author
Johanna Dubos, MSc
From the Institut de Bio-imagerie IBIO (F.M., J.D., V.D., T.T.), University Bordeaux; Neuroimagerie Diagnostique et Thérapeutique (D.P., G.M., T.C., V.D., T.T.), CHU de Bordeaux, France; Kansai Electric Power Hospital (H.F.), Osaka, Japan; Inserm CIC-IT U1433 (E.M., B.C.), CHRU Nancy; Institut de Psychiatrie et Neurosciences de Paris (IPNP) (P.S.), INSERM U1266; Département de Neurologie (P.S.), Hopital Fondation Rothschild, Paris; Institut des Maladies Neurodégénératives (V.P.), CNRS, UMR 5293, Bordeaux INP (P.C.), LABRI, CNRS, UMR5800, and Neurocentre Magendie (V.D., T.T.), INSERM U1215, Univ. Bordeaux; Service de Neurologie et Unité de Neuro Vasculaire (B.L.), Hôpital FOCH, Suresnes; Unité Neurovasculaire (J.M.O.), CHU de Toulouse; Unité Neurovasculaire (I.S.), CHU de Bordeaux; CNRS (M.T.D.S.), UMR-5293, Univ. Bordeaux; and Brain Connectivity and Behaviour Laboratory (M.T.D.S.), Paris, France.
Disclosure
Financial Disclosure:
1.
NONE
Research Support:
1.
NONE
Stock, Stock Options & Royalties:
1.
NONE
Legal Proceedings:
1.
NONE
View all articles by this author
Vincent Planche, MD, PhD https://orcid.org/0000-0003-3713-227X
From the Institut de Bio-imagerie IBIO (F.M., J.D., V.D., T.T.), University Bordeaux; Neuroimagerie Diagnostique et Thérapeutique (D.P., G.M., T.C., V.D., T.T.), CHU de Bordeaux, France; Kansai Electric Power Hospital (H.F.), Osaka, Japan; Inserm CIC-IT U1433 (E.M., B.C.), CHRU Nancy; Institut de Psychiatrie et Neurosciences de Paris (IPNP) (P.S.), INSERM U1266; Département de Neurologie (P.S.), Hopital Fondation Rothschild, Paris; Institut des Maladies Neurodégénératives (V.P.), CNRS, UMR 5293, Bordeaux INP (P.C.), LABRI, CNRS, UMR5800, and Neurocentre Magendie (V.D., T.T.), INSERM U1215, Univ. Bordeaux; Service de Neurologie et Unité de Neuro Vasculaire (B.L.), Hôpital FOCH, Suresnes; Unité Neurovasculaire (J.M.O.), CHU de Toulouse; Unité Neurovasculaire (I.S.), CHU de Bordeaux; CNRS (M.T.D.S.), UMR-5293, Univ. Bordeaux; and Brain Connectivity and Behaviour Laboratory (M.T.D.S.), Paris, France.
Disclosure
Financial Disclosure:
1.
Personal Compensation: (1) Consulting Fees for Animal Studies (MRI) - Motac Neuroscience Ltd
Research Support:
1.
(1) Foundation - Fondation PSP France (NA): Preclinical/Animal Studies (2) Governmental - Center of Excellence for Neurodegenerative Diseases (NA): Preclinical/Animal Studies (3) Foundation - Fondation Recherche Alzheimer (NA): Preclinical/Animal Studies
Stock, Stock Options & Royalties:
1.
NONE
Legal Proceedings:
1.
NONE
View all articles by this author
Pierrick Coupé, PhD https://orcid.org/0000-0003-2709-3350
From the Institut de Bio-imagerie IBIO (F.M., J.D., V.D., T.T.), University Bordeaux; Neuroimagerie Diagnostique et Thérapeutique (D.P., G.M., T.C., V.D., T.T.), CHU de Bordeaux, France; Kansai Electric Power Hospital (H.F.), Osaka, Japan; Inserm CIC-IT U1433 (E.M., B.C.), CHRU Nancy; Institut de Psychiatrie et Neurosciences de Paris (IPNP) (P.S.), INSERM U1266; Département de Neurologie (P.S.), Hopital Fondation Rothschild, Paris; Institut des Maladies Neurodégénératives (V.P.), CNRS, UMR 5293, Bordeaux INP (P.C.), LABRI, CNRS, UMR5800, and Neurocentre Magendie (V.D., T.T.), INSERM U1215, Univ. Bordeaux; Service de Neurologie et Unité de Neuro Vasculaire (B.L.), Hôpital FOCH, Suresnes; Unité Neurovasculaire (J.M.O.), CHU de Toulouse; Unité Neurovasculaire (I.S.), CHU de Bordeaux; CNRS (M.T.D.S.), UMR-5293, Univ. Bordeaux; and Brain Connectivity and Behaviour Laboratory (M.T.D.S.), Paris, France.
Disclosure
Financial Disclosure:
1.
NONE
Research Support:
1.
NONE
Stock, Stock Options & Royalties:
1.
NONE
Legal Proceedings:
1.
NONE
View all articles by this author
Vincent Dousset, MD, PhD https://orcid.org/0000-0003-0427-3236
From the Institut de Bio-imagerie IBIO (F.M., J.D., V.D., T.T.), University Bordeaux; Neuroimagerie Diagnostique et Thérapeutique (D.P., G.M., T.C., V.D., T.T.), CHU de Bordeaux, France; Kansai Electric Power Hospital (H.F.), Osaka, Japan; Inserm CIC-IT U1433 (E.M., B.C.), CHRU Nancy; Institut de Psychiatrie et Neurosciences de Paris (IPNP) (P.S.), INSERM U1266; Département de Neurologie (P.S.), Hopital Fondation Rothschild, Paris; Institut des Maladies Neurodégénératives (V.P.), CNRS, UMR 5293, Bordeaux INP (P.C.), LABRI, CNRS, UMR5800, and Neurocentre Magendie (V.D., T.T.), INSERM U1215, Univ. Bordeaux; Service de Neurologie et Unité de Neuro Vasculaire (B.L.), Hôpital FOCH, Suresnes; Unité Neurovasculaire (J.M.O.), CHU de Toulouse; Unité Neurovasculaire (I.S.), CHU de Bordeaux; CNRS (M.T.D.S.), UMR-5293, Univ. Bordeaux; and Brain Connectivity and Behaviour Laboratory (M.T.D.S.), Paris, France.
Disclosure
Financial Disclosure:
1.
NONE
Research Support:
1.
NONE
Stock, Stock Options & Royalties:
1.
NONE
Legal Proceedings:
1.
NONE
View all articles by this author
Bertrand Lapergue, MD, PhD https://orcid.org/0000-0002-8993-2175
From the Institut de Bio-imagerie IBIO (F.M., J.D., V.D., T.T.), University Bordeaux; Neuroimagerie Diagnostique et Thérapeutique (D.P., G.M., T.C., V.D., T.T.), CHU de Bordeaux, France; Kansai Electric Power Hospital (H.F.), Osaka, Japan; Inserm CIC-IT U1433 (E.M., B.C.), CHRU Nancy; Institut de Psychiatrie et Neurosciences de Paris (IPNP) (P.S.), INSERM U1266; Département de Neurologie (P.S.), Hopital Fondation Rothschild, Paris; Institut des Maladies Neurodégénératives (V.P.), CNRS, UMR 5293, Bordeaux INP (P.C.), LABRI, CNRS, UMR5800, and Neurocentre Magendie (V.D., T.T.), INSERM U1215, Univ. Bordeaux; Service de Neurologie et Unité de Neuro Vasculaire (B.L.), Hôpital FOCH, Suresnes; Unité Neurovasculaire (J.M.O.), CHU de Toulouse; Unité Neurovasculaire (I.S.), CHU de Bordeaux; CNRS (M.T.D.S.), UMR-5293, Univ. Bordeaux; and Brain Connectivity and Behaviour Laboratory (M.T.D.S.), Paris, France.
Disclosure
Financial Disclosure:
1.
NONE
Research Support:
1.
NONE
Stock, Stock Options & Royalties:
1.
NONE
Legal Proceedings:
1.
NONE
View all articles by this author
Jean Marc Olivot, MD, PhD https://orcid.org/0000-0003-2027-2276
From the Institut de Bio-imagerie IBIO (F.M., J.D., V.D., T.T.), University Bordeaux; Neuroimagerie Diagnostique et Thérapeutique (D.P., G.M., T.C., V.D., T.T.), CHU de Bordeaux, France; Kansai Electric Power Hospital (H.F.), Osaka, Japan; Inserm CIC-IT U1433 (E.M., B.C.), CHRU Nancy; Institut de Psychiatrie et Neurosciences de Paris (IPNP) (P.S.), INSERM U1266; Département de Neurologie (P.S.), Hopital Fondation Rothschild, Paris; Institut des Maladies Neurodégénératives (V.P.), CNRS, UMR 5293, Bordeaux INP (P.C.), LABRI, CNRS, UMR5800, and Neurocentre Magendie (V.D., T.T.), INSERM U1215, Univ. Bordeaux; Service de Neurologie et Unité de Neuro Vasculaire (B.L.), Hôpital FOCH, Suresnes; Unité Neurovasculaire (J.M.O.), CHU de Toulouse; Unité Neurovasculaire (I.S.), CHU de Bordeaux; CNRS (M.T.D.S.), UMR-5293, Univ. Bordeaux; and Brain Connectivity and Behaviour Laboratory (M.T.D.S.), Paris, France.
Disclosure
Financial Disclosure:
1.
Personal Compensation: (1) Consulting - Abbvie (2) Consulting - Boehringer-Ingelheim (3) Consulting - Acticor (4) Speaker fee - Bristol Myers Squibb
Research Support:
1.
(1) French Ministry of Health - PHRC and ANR: RCT
Stock, Stock Options & Royalties:
1.
NONE
Legal Proceedings:
1.
NONE
View all articles by this author
Igor Sibon, MD, PhD https://orcid.org/0000-0002-1171-4215
From the Institut de Bio-imagerie IBIO (F.M., J.D., V.D., T.T.), University Bordeaux; Neuroimagerie Diagnostique et Thérapeutique (D.P., G.M., T.C., V.D., T.T.), CHU de Bordeaux, France; Kansai Electric Power Hospital (H.F.), Osaka, Japan; Inserm CIC-IT U1433 (E.M., B.C.), CHRU Nancy; Institut de Psychiatrie et Neurosciences de Paris (IPNP) (P.S.), INSERM U1266; Département de Neurologie (P.S.), Hopital Fondation Rothschild, Paris; Institut des Maladies Neurodégénératives (V.P.), CNRS, UMR 5293, Bordeaux INP (P.C.), LABRI, CNRS, UMR5800, and Neurocentre Magendie (V.D., T.T.), INSERM U1215, Univ. Bordeaux; Service de Neurologie et Unité de Neuro Vasculaire (B.L.), Hôpital FOCH, Suresnes; Unité Neurovasculaire (J.M.O.), CHU de Toulouse; Unité Neurovasculaire (I.S.), CHU de Bordeaux; CNRS (M.T.D.S.), UMR-5293, Univ. Bordeaux; and Brain Connectivity and Behaviour Laboratory (M.T.D.S.), Paris, France.
Disclosure
Financial Disclosure:
1.
NONE
Research Support:
1.
NONE
Stock, Stock Options & Royalties:
1.
NONE
Legal Proceedings:
1.
(1) Consultant for Novonordisk in 2023 (2) Consultant for Novartis in 2023 (3) Consultant for Medtronic in 2022 (4) Consultant for Sanofi in 2022 (5) Consultant for BMS-Pfizer in 2022 (6) Consultant for Boheringer Ingelheim in 2022
View all articles by this author
Michel Thiebaut De Schotten, PhD https://orcid.org/0000-0002-0329-1814
From the Institut de Bio-imagerie IBIO (F.M., J.D., V.D., T.T.), University Bordeaux; Neuroimagerie Diagnostique et Thérapeutique (D.P., G.M., T.C., V.D., T.T.), CHU de Bordeaux, France; Kansai Electric Power Hospital (H.F.), Osaka, Japan; Inserm CIC-IT U1433 (E.M., B.C.), CHRU Nancy; Institut de Psychiatrie et Neurosciences de Paris (IPNP) (P.S.), INSERM U1266; Département de Neurologie (P.S.), Hopital Fondation Rothschild, Paris; Institut des Maladies Neurodégénératives (V.P.), CNRS, UMR 5293, Bordeaux INP (P.C.), LABRI, CNRS, UMR5800, and Neurocentre Magendie (V.D., T.T.), INSERM U1215, Univ. Bordeaux; Service de Neurologie et Unité de Neuro Vasculaire (B.L.), Hôpital FOCH, Suresnes; Unité Neurovasculaire (J.M.O.), CHU de Toulouse; Unité Neurovasculaire (I.S.), CHU de Bordeaux; CNRS (M.T.D.S.), UMR-5293, Univ. Bordeaux; and Brain Connectivity and Behaviour Laboratory (M.T.D.S.), Paris, France.
Disclosure
Financial Disclosure:
1.
NONE
Research Support:
1.
NONE
Stock, Stock Options & Royalties:
1.
NONE
Legal Proceedings:
1.
NONE
View all articles by this author
Thomas Tourdias, MD, PhD https://orcid.org/0000-0002-7151-6325
From the Institut de Bio-imagerie IBIO (F.M., J.D., V.D., T.T.), University Bordeaux; Neuroimagerie Diagnostique et Thérapeutique (D.P., G.M., T.C., V.D., T.T.), CHU de Bordeaux, France; Kansai Electric Power Hospital (H.F.), Osaka, Japan; Inserm CIC-IT U1433 (E.M., B.C.), CHRU Nancy; Institut de Psychiatrie et Neurosciences de Paris (IPNP) (P.S.), INSERM U1266; Département de Neurologie (P.S.), Hopital Fondation Rothschild, Paris; Institut des Maladies Neurodégénératives (V.P.), CNRS, UMR 5293, Bordeaux INP (P.C.), LABRI, CNRS, UMR5800, and Neurocentre Magendie (V.D., T.T.), INSERM U1215, Univ. Bordeaux; Service de Neurologie et Unité de Neuro Vasculaire (B.L.), Hôpital FOCH, Suresnes; Unité Neurovasculaire (J.M.O.), CHU de Toulouse; Unité Neurovasculaire (I.S.), CHU de Bordeaux; CNRS (M.T.D.S.), UMR-5293, Univ. Bordeaux; and Brain Connectivity and Behaviour Laboratory (M.T.D.S.), Paris, France.
Disclosure
Financial Disclosure:
1.
NONE
Research Support:
1.
(1) ANR - TRAIL (ANR-10-LABX-57): Research program on biomedical imaging (2) ANR - IMPACT (ANR-10-IDEX-03-02): Research program on biomedical imaging
Stock, Stock Options & Royalties:
1.
NONE
Legal Proceedings:
1.
NONE
View all articles by this author
for the FRAME and ETIS investigators
From the Institut de Bio-imagerie IBIO (F.M., J.D., V.D., T.T.), University Bordeaux; Neuroimagerie Diagnostique et Thérapeutique (D.P., G.M., T.C., V.D., T.T.), CHU de Bordeaux, France; Kansai Electric Power Hospital (H.F.), Osaka, Japan; Inserm CIC-IT U1433 (E.M., B.C.), CHRU Nancy; Institut de Psychiatrie et Neurosciences de Paris (IPNP) (P.S.), INSERM U1266; Département de Neurologie (P.S.), Hopital Fondation Rothschild, Paris; Institut des Maladies Neurodégénératives (V.P.), CNRS, UMR 5293, Bordeaux INP (P.C.), LABRI, CNRS, UMR5800, and Neurocentre Magendie (V.D., T.T.), INSERM U1215, Univ. Bordeaux; Service de Neurologie et Unité de Neuro Vasculaire (B.L.), Hôpital FOCH, Suresnes; Unité Neurovasculaire (J.M.O.), CHU de Toulouse; Unité Neurovasculaire (I.S.), CHU de Bordeaux; CNRS (M.T.D.S.), UMR-5293, Univ. Bordeaux; and Brain Connectivity and Behaviour Laboratory (M.T.D.S.), Paris, France.
View all articles by this author

Notes

Correspondence Dr. Munsch [email protected]
Go to Neurology.org/N for full disclosures. Funding information and disclosures deemed relevant by the authors, if any, are provided at the end of the article.
Coinvestigators are listed at Neurology.org.
Previously published at bioRxiv doi: 10.1101/2023.12.12.571221.
Submitted and externally peer reviewed. The handling editor was Editor-in-Chief José Merino, MD, MPhil, FAAN.
*
These authors contributed equally to this work as cofirst authors.

Metrics & Citations

Metrics

Citation information is sourced from Crossref Cited-by service.

Citations

Download Citations

If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Select your manager software from the list below and click Download.

Cited By
    • Umberto Pensato,
    • Andrew M. Demchuk,
    • Bijoy K. Menon,
    • Thanh N. Nguyen,
    • Gabriel Broocks,
    • Bruce C.V. Campbell,
    • Diego A. Gutierrez Vasquez,
    • Peter J. Mitchell,
    • Michael D. Hill,
    • Mayank Goyal,
    • Johanna M. Ospel,
    Cerebral Infarct Growth: Pathophysiology, Pragmatic Assessment, and Clinical Implications, Stroke, 56, 1, (219-229), (2025).https://doi.org/10.1161/STROKEAHA.124.049013
    Crossref
  2. Dynamic Evolution of Infarct Volumes at MRI in Ischemic Stroke Due to Large Vessel Occlusion, Neurology, 103, 5, (2024)./doi/10.1212/WNL.0000000000209782
    Abstract

View Options

View options

PDF and All Supplements

Download PDF and Supplementary Material

Short Form

Short Form

Full Text

View Full Text
Login options

Check if you have access through your login credentials or your institution to get full access on this article.

Personal login Institutional Login
Purchase Options

The neurology.org payment platform is currently offline. Our technical team is working as quickly as possible to restore service.

If you need immediate support or to place an order, please call or email customer service:

  • 1-800-638-3030 for U.S. customers - 8:30 - 7 pm ET (M-F)
  • 1-301-223-2300 for customers outside the U.S. - 8:30 - 7 pm ET (M-F)
  • [email protected]

We appreciate your patience during this time and apologize for any inconvenience.

Figures

Tables

Media

Share

Share

Share article link

Share

View full text|Download PDF