Skip main navigation

ASCOD Phenotyping of Stroke With Anterior Large Vessel Occlusion Treated by Mechanical Thrombectomy

and on behalf of the FRAME Investigators*
Originally published 2021;52:e769–e772


Background and Purpose:

Determining the mechanism of large vessel occlusion related acute ischemic stroke is of major importance to initiate a tailored secondary prevention strategy. We investigated using the atherosclerosis, small vessel disease, cardiac source, other cause, dissection (ASCOD) classification the distribution of the causes of large vessel occlusion related acute ischemic stroke treated by mechanical thrombectomy.


This was a predefined substudy of the FRAME (French Acute Multimodal Imaging to Select Patient for Mechanical Thrombectomy). Each patient underwent a systematic etiological workup including brain and vascular imaging, electrocardiogram monitoring lasting at least 24 hours and routine blood tests. Stroke mechanisms were systematically evaluated using the atherosclerosis, small vessel disease, cardiac source, other cause, dissection grading system at 3 months. We defined single potential cause by one cause graded 1 in a single domain, possible cause as a cause graded 1 or 2 regardless of overlap, and no identified cause without grade 1 nor 2 causes.


A total of 215 patients (mean age 70±14; 50% male) were included. A single potential cause was identified in 148 (69%). Cardio-embolism (53%) was the most frequent, followed by atherosclerosis (9%), dissection (5%) and other causes (1%). Atrial fibrillation accounted for 88% of C1. Overlap between grade 1 causes was uncommon (3%). Possible causes were identified in 168 patients (83%) and 16 (7%) had no cause identified after the initial evaluation.


Cardio-embolism, especially atrial fibrillation, was the major cause of large vessel occlusion related acute ischemic stroke. This finding emphasizes the yield of paroxysmal atrial fibrillation detection in those patients.


URL:; Unique identifier: NCT03045146.

Determining the underlying cause(s) of large vessel occlusion related acute ischemic stroke (LVO-AIS) is essential to initiate promptly a tailored secondary prevention.

Previous studies have suggested using TOAST (Trial of ORG 10172 in Acute Stroke Treatment) classification that cardio-embolism is a major cause of LVO-AIS.1 However, those studies were not designed to describe stroke causes. Therefore, etiological workup could remain uncompleted once a single cause was identified, neglecting the fact that several causes may coexist. A systematic prospective evaluation using the atherosclerosis, small vessel disease, cardiac source, other cause, dissection (ASCOD) grading has not been performed. Compared with TOAST, ASCOD2,3 has the advantage to assess the overlap of several mechanisms and evaluate the completeness of the etiological work-up.

The aim of this study is to describe, in a prospective cohort of patients treated by mechanical thrombectomy (MT), the distribution of LVO-AIS causes using a systematic ASCOD classification.


This is a predefined FRAME (French Acute Multimodal Imaging to Select Patient for Mechanical Thrombectomy)4 substudy. FRAME is a prospective cohort study of patients experiencing anterior LVO-AIS treated by MT within 6 hours after onset. Each patient underwent a systematic etiological evaluation. Eligibility criteria and etiological workup are summarized in the supplemental material.

The primary outcome was the distribution of the different causes, according to the ASCOD classification3 prospectively determined 3 months after MT.

We assessed the prevalence of:

  1. Single potential cause defined by a cause graded 1 in a single domain, without any other category graded 1.

  2. Possible cause defined by at least one cause graded 1 or 2.

  3. Overlap between grade 1 and between grade 1–2 causes

  4. No identified cause, defined by:

    • – Negative minimal work up, no cause graded 1 or 2, nor incomplete workup (no grade 9).

    • – Incomplete work-up, no cause graded 1 or 2, and at least one incomplete work up (grade 9).

The trial protocol was approved by the French Ethical Committee (CPP SOOM III) on October 5, 2016 and was authorized by the French Health Authority. Every patient or his/her legal representative signed a written informed consent at inclusion. The data are available from the corresponding authors on reasonable request. STROBE checklist and flow diagram can be found in the Supplemental Material.


Two-hundred and fifteen patients were enrolled in this substudy. Baseline characteristics are summarized in Table 1. ASCOD grading distribution is available in Table I in the Supplemental Material. Stroke subtype distribution is summarized on Table 2. One hundred fifteen (53%) patients had a single potential cardio-embolic cause (C1), and 134 (62%) C1 or C2. Among the 121 patients graded C1 (including 115 isolated C1 cause), 106 (88%) had atrial fibrillation (AF): 43 (20%) patients had history of AF, 31 (14%) had AF detected on admission, 21 (10%) during the rest of hospitalization, and 11 (5%) after discharge. Among the 90 (42%) patients with no potential arterial cause such as atherosclerosis or dissection (ie, A1–A2 or D1–D2), and no history of AF or detection on admission; 29 (32%) had a diagnosis of paroxysmal AF (pAF) at 3 months including 19 (22%) during hospitalization.

Table 1. Baseline Characteristics

Baseline characteristicsAll patients, N=215
Age, mean (SD)70 (14)
Gender female, n (%)107 (50)
Baseline NIHSS, median (IQR)17 (11–21)
Hypertension, n (%)125 (58)
Diabetes, n (%)33 (15)
Hyperlipidemia, n (%)59 (27)
Active smoker, n (%)34 (16)
Peripheral vascular disease, n (%)4 (2)
Previous myocardial infarction, n (%)13 (6)
Previous transient ischemic attack, n (%)11 (5)
Previous ischemic stroke, n (%)23 (11)
History of atrial fibrillation, n (%)43 (20)
Anticoagulation, n (%)33 (15)
Antiplatelet therapy, n (%)56 (26)
Brain imaging MRI/CT, n (%)205/10 (95/5)
Vascular imaging, n (%)209 (97)

Data are expressed as mean (SD), median (IQR), or n (%). CT indicates computerized tomography; IQR, interquartile range; MRI, magnetic resonance imaging; and NIHSS, National Institutes of Health Stroke Scale.

Table 2. Stroke Subtype Distribution

Single cause (isolated grade 1), n (%)148 (69%)
 C1, n (%)115 (53%)
 A1, n (%)20 (9%)
 O1, n (%)3 (1%)
 D1, n (%)10 (5%)
Possible cause (at least one cause grade 1 or 2), n (%)178 (83%)
 C1/C2, n (%)134 (62%)
 A1/A2, n (%)49 (23%)
 O1/O2, n (%)7 (3%)
 D1/D2, n (%)10 (5%)
No identified cause (grade 3, 0, or 9), n (%)37 (17%)
 Incomplete minimal work up21(10%)
 Negative complete minimal work up16 (7%)

Six (3%) patients had coexisting causes graded 1: Two patients were C1-A1 and 4 C1-O1 (Figure [A]). Twenty-one (10%) patients had overlapping causes graded 1 or 2: 17 (8%) had cardio-embolism and atherosclerosis, 3 (1%) had cardio-embolism and other cause, and 1 (0.5%) patient cumulated cardio-embolism, atherosclerosis, and other cause. Eleven (5%) patients had coexisting causes graded C1 and A2, representing the most frequent phenotype of stroke with overlapping causes (Figure [B]).


Figure. Overlap between atherosclerosis, cardio-embolism, and dissection mechanisms. Venn diagram: (A) overlap between grade 1 and (B) overlap between grade 1 or 2.

Stroke cause remained undetermined in 37 (17%) patients: 16 (7%) had a negative minimal work-up. Twenty-one (10%) had an incomplete work-up, 18 (8%) were graded C9 (ie, incomplete cardiac work-up).


Our results confirm that, using systematic ASCOD grading at 3 months, cardio-embolism, especially AF, is the most common cause of LVO-AIS. In addition, overlap and undetermined mechanisms are uncommon.

More than half of LVO-AIS were caused by cardio-embolism, largely represented by AF. Furthermore, prevalence of AF is probably underestimated, due to the rate of incomplete initial cardiac exploration (8%) and the absence of systematic prolonged electrocardiogram monitoring after discharge in the majority of patients with no identified cause.

This remarkably high prevalence of cardio-embolism seems to be characteristic of LVO-AIS by comparison with an unselected AIS population, where the rate of grade C1 is about 35% after the exclusion of lacunar strokes.5,6 This finding may be explained by large clot size in AF; these larger clots are more likely to occlude proximal cerebral arteries.7

Paroxysmal AF was detected during hospitalization among one-fourth of the 90 patients without arterial cause or AF on admission (no history of AF nor AF detected on admission). This high incidence contrasts with a previously reported 5% rate among unselected AIS population.8 This finding suggests that early detection of pAF after LVO-AIS is common and confirms the benefits of cardiac monitoring in these patients.7

In our study, AIS without identified cause was, using either ASCOD (17%) or TOAST (23%) classifications, lower than previously reported using TOAST (35%).1 This is potentially explained by the prospective design of our study that involves systematic follow-up and data collection, potentially limiting the rate of incomplete evaluation (10%).

Overlapping causes graded 1 were overall uncommon in our cohort. Nonetheless, as previously described5 we report that underlying pAF can be detected in patients with moderate atherosclerosis (graded A2), emphasizing the need for systematic screening for both causes after LVO-AIS.

The relatively high rate of dissection among LVO-AIS has been previously reproted.9 This can also been explained by the fact that patient with arterial dissection are younger, therefore may be more frequently addressed to MT compared with elderly and comorbid patient.

Our study suffers from some limitations: First, its small sample size. Second, the exclusion of patients with past history of debilitating stroke or underlying diseases limiting their life expectancy may have led to selection bias. Third, ASCOD classification was evaluated after 90 days which may not have allowed the completion of an exhaustive work-up (eg, implantation of insertable cardiac monitoring).

The vast majority of FRAME study participants only has short-term electrocardiogram monitoring, which is time consuming and less efficient that long-term electrocardiogram using insertable cardiac monitoring to detect pAF.10 A systematic, early, and intensive screening strategies, using insertable cardiac monitoring implantation before discharge, could be particularly cost-effective in this subgroup of patients.


Cardio-embolism, mostly AF, was the major cause of LVO-AIS treated by MT.

Prospective studies aiming to detect pAF using intensive screening strategies such as early insertable cardiac monitoring implantation in patients with LVO-AIS with no identified cause is warranted.

Article Information

Supplemental Materials

Expanded Materials and Methods

Supplemental Figures I and II

Supplemental Tables I and II

Nonstandard Abbreviations and Acronyms


atrial fibrillation


acute ischemic stroke


atherosclerosis, small vessel disease, cardiac source, other cause, dissection


French Acute Multimodal Imaging to Select Patient for Mechanical Thrombectomy


large vessel occlusion


mechanical thrombectomy


paroxysmal atrial fibrillation


Liste of FRAME Investigators

Soren Christensen, Stanford Stroke Center, Stanford, CA; Michael Mlynash, Stanford Stroke Center, Stanford, CA; François Chollet, Neurology, CHU Toulouse, France; Marianne Barbieux, Neurology, CHU Toulouse, France; Caterina Michelozzi, Neuroradiology, CHU Toulouse, France; Philippe Tall, Neuroradiology, CHU Toulouse, France; François Caparros, Neurology, CHU Lille, France; Brigitte Pouzet, Clinical Investigation Center, CHU Toulouse, France; Fabienne Calvas, Clinical Investigation Center, CHU Toulouse, France; Monique Galitzki, Clinical Investigation Center, CHU Toulouse, France; Pauline Renou, Neurology, CHU Bordeaux, France; François Rouanet, Neurology, CHU Bordeaux, France; Jerome Berge, Neuroradiology, CHU Bordeaux, France; Gauthier Marnat, Neuroradiology, CHU Bordeaux, France; Ludovic Lucas, Neurology, CHU Bordeaux, France; Cyrielle Coignon, Neurology, CHU Bordeaux, France; Sharmila Sagnier, Neurology, CHU Bordeaux, France; Sabrina Debruxelle, Neurology, CHU Bordeaux, France; Sylvain Ledure, Neurology, CHU Bordeaux, France.


*A list of the FRAME investigators is available in the Appendix.

Supplemental Material is available with this article at

For Sources of Funding and Disclosures, see page e772.

Presented in part at the European Stroke Organisation-World Stroke Organization 2020 Virtual Conference, November 7–9, 2020.

Correspondence to: Louis Fontaine, MD, Acute Stroke Unit, Toulouse University Hospital, 1 Place du Dr Baylac, 31059 Toulouse, France. Email


  • 1. Meinel TR, Kaesmacher J, Mordasini P, Mosimann PJ, Jung S, Arnold M, Heldner MR, Michel P, Hajdu SD, Ribo M, et al. Outcome, efficacy and safety of endovascular thrombectomy in ischaemic stroke according to time to reperfusion: data from a multicentre registry.Ther Adv Neurol Disord. 2019; 12:1756286419835708. doi: 10.1177/1756286419835708CrossrefGoogle Scholar
  • 2. Amarenco P, Bogousslavsky J, Caplan LR, Donnan GA, Hennerici MG. New approach to stroke subtyping: the A-S-C-O (phenotypic) classification of stroke.Cerebrovasc Dis. 2009; 27:502–508. doi: 10.1159/000210433CrossrefMedlineGoogle Scholar
  • 3. Amarenco P, Bogousslavsky J, Caplan LR, Donnan GA, Wolf ME, Hennerici MG. The ASCOD phenotyping of ischemic stroke (Updated ASCO Phenotyping).Cerebrovasc Dis. 2013; 36:1–5. doi: 10.1159/000352050CrossrefMedlineGoogle Scholar
  • 4. Olivot JM, Albucher JF, Guenego A, Thalamas C, Mlynash M, Rousseau V, Drif A, Christensen S, Sommet A, Viguier A, et al; FRAME Investigators*. Mismatch profile influences outcome after mechanical Thrombectomy.Stroke. 2021; 52:232–240. doi: 10.1161/STROKEAHA.120.031929LinkGoogle Scholar
  • 5. Sirimarco G, Lavallée PC, Labreuche J, Meseguer E, Cabrejo L, Guidoux C, Klein IF, Olivot JM, Abboud H, Adraï V, et al. Overlap of diseases underlying ischemic stroke: the ASCOD phenotyping.Stroke. 2013; 44:2427–2433. doi: 10.1161/STROKEAHA.113.001363LinkGoogle Scholar
  • 6. Wolf ME, Sauer T, Alonso A, Hennerici MG. Comparison of the new ASCO classification with the TOAST classification in a population with acute ischemic stroke.J Neurol. 2012; 259:1284–1289. doi: 10.1007/s00415-011-6325-1CrossrefMedlineGoogle Scholar
  • 7. Pagola J, Juega J, Francisco-Pascual J, Bustamante A, Penalba A, Pala E, Rodriguez M, De Lera Alfonso M, Arenillas JF, Cabezas JA, et al; Crypto-AF study group. Large vessel occlusion is independently associated with atrial fibrillation detection.Eur J Neurol. 2020; 27:1618–1624. doi: 10.1111/ene.14281CrossrefMedlineGoogle Scholar
  • 8. Sposato LA, Cipriano LE, Saposnik G, Ruíz Vargas E, Riccio PM, Hachinski V. Diagnosis of atrial fibrillation after stroke and transient ischaemic attack: a systematic review and meta-analysis.Lancet Neurol. 2015; 14:377–387. doi: 10.1016/S1474-4422(15)70027-XCrossrefMedlineGoogle Scholar
  • 9. Compagne KCJ, Goldhoorn RB, Uyttenboogaart M, van Oostenbrugge RJ, van Zwam WH, van Doormaal PJ, Dippel DWJ, van der Lugt A, Emmer BJ, van Es ACGM; MR CLEAN investigators. Acute endovascular treatment of patients with Ischemic Stroke from intracranial large vessel occlusion and extracranial carotid dissection.Front Neurol. 2019; 10:102. doi: 10.3389/fneur.2019.00102CrossrefGoogle Scholar
  • 10. Sanna T, Diener HC, Passman RS, Di Lazzaro V, Bernstein RA, Morillo CA, Rymer MM, Thijs V, Rogers T, Beckers F, et al; CRYSTAL AF Investigators. Cryptogenic stroke and underlying atrial fibrillation.N Engl J Med. 2014; 370:2478–2486. doi: 10.1056/NEJMoa1313600CrossrefMedlineGoogle Scholar


eLetters should relate to an article recently published in the journal and are not a forum for providing unpublished data. Comments are reviewed for appropriate use of tone and language. Comments are not peer-reviewed. Acceptable comments are posted to the journal website only. Comments are not published in an issue and are not indexed in PubMed. Comments should be no longer than 500 words and will only be posted online. References are limited to 10. Authors of the article cited in the comment will be invited to reply, as appropriate.

Comments and feedback on AHA/ASA Scientific Statements and Guidelines should be directed to the AHA/ASA Manuscript Oversight Committee via its Correspondence page.