Journal of Neurology (2020) 267:649–658 https://doi.org/10.1007/s00415-019-09613-5 ORIGINAL COMMUNICATION Detailed phenotyping of posterior vs. anterior circulation ischemic stroke: a multi‑center MRI study Petrea Frid1,2,3 · Mattias Drake4,5 · A. K. Giese6,7 · J. Wasselius4,5 · M. D. Schirmer6,8,9 · K. L. Donahue6 · L. Cloonan6 · R. Irie10 · M. J. R. J. Bouts10 · E. C. McIntosh10 · S. J. T. Mocking10 · A. V. Dalca8,10 · R. Sridharan8 · H. Xu11 · E. Giralt‑Steinhauer12 · L. Holmegaard13,14 · K. Jood13,14 · J. Roquer12 · J. W. Cole15 · P. F. McArdle11 · J. P. Broderick16 · J. Jimenez‑Conde12 · C. Jern17 · B. M. Kissela16 · D. O. Kleindorfer16 · R. Lemmens18,19,20 · J. F. Meschia21 · T. Rundek22 · R. L. Sacco22 · R. Schmidt23 · P. Sharma24,25 · A. Slowik26 · V. Thijs27,28 · D. Woo16 · B. B. Worrall29,30 · S. J. Kittner15 · B. D. Mitchell11,31 · J. Petersson1,2 · J. Rosand6,7,10,32,33 · P. Golland8 · O. Wu10 · N. S. Rost6 · A. Lindgren1,34 on behalf of the Stroke Genetics Network (SiGN), the International Stroke Genetics Consortium (ISGC), and the MRI-Genetics Interface Exploration (MRI-GENIE) Study Received: 4 September 2019 / Revised: 28 October 2019 / Accepted: 30 October 2019 / Published online: 11 November 2019 © The Author(s) 2019 Abstract Objective Posterior circulation ischemic stroke (PCiS) constitutes 20–30% of ischemic stroke cases. Detailed information about differences between PCiS and anterior circulation ischemic stroke (ACiS) remains scarce. Such information might guide clinical decision making and prevention strategies. We studied risk factors and ischemic stroke subtypes in PCiS vs. ACiS and lesion location on magnetic resonance imaging (MRI) in PCiS. Methods Out of 3,301 MRIs from 12 sites in the National Institute of Neurological Disorders and Stroke (NINDS) Stroke Genetics Network (SiGN), we included 2,381 cases with acute DWI lesions. The definition of ACiS or PCiS was based on lesion location. We compared the groups using Chi-squared and logistic regression. Results PCiS occurred in 718 (30%) patients and ACiS in 1663 (70%). Diabetes and male sex were more common in PCiS vs. ACiS (diabetes 27% vs. 23%, p < 0.05; male sex 68% vs. 58%, p < 0.001). Both were independently associated with PCiS (diabetes, OR = 1.29; 95% CI 1.04–1.61; male sex, OR = 1.46; 95% CI 1.21–1.78). ACiS more commonly had large artery atherosclerosis (25% vs. 20%, p < 0.01) and cardioembolic mechanisms (17% vs. 11%, p < 0.001) compared to PCiS. Small artery occlusion was more common in PCiS vs. ACiS (20% vs. 14%, p < 0.001). Small artery occlusion accounted for 47% of solitary brainstem infarctions. Conclusion Ischemic stroke subtypes differ between the two phenotypes. Diabetes and male sex have a stronger association with PCiS than ACiS. Definitive MRI-based PCiS diagnosis aids etiological investigation and contributes additional insights into specific risk factors and mechanisms of injury in PCiS. Keywords Stroke · Posterior circulation brain infarction · Risk factors · Magnetic resonance imaging · Phenotyping Introduction Posterior circulation ischemic stroke (PCiS) refers to infarc- tion in any brain structure supplied by the vertebrobasilar arterial system. The reported prevalence of PCiS ranges Electronic supplementary material The online version of this between 20 and 30% in most hospital-based cohorts [1–3]. article (https​://doi.org/10.1007/s0041​5-019-09613​-5) contains To a large extent, PCiS and anterior circulation ischemic supplementary material, which is available to authorized users. stroke (ACiS) share vascular risk factors and stroke mecha- nisms [1, 4]. PCiS has been reported to occur more com- * Petrea Frid petrea.frid@med.lu.se monly than ACiS in young patients in whom cervical artery dissection is more frequent [5] and in patients with MELAS Extended author information available on the last page of the article 13 Vol.:(0123456789) 650 Journal of Neurology (2020) 267:649–658 and Fabry’s disease [6]. Previous studies comparing PCiS age, sex, hypertension, diabetes mellitus, atrial fibrillation, and ACiS suffer from heterogeneous sample sizes, diag- smoking status, and ischemic stroke mechanism. Ten sites nostic criteria, imaging methods, and subtype classifica- contributed data on stroke severity according to the National tion. The reported prevalence of conventional vascular risk Institute of Health Stroke Scale (NIHSS). MRI lesion loca- factors such as hypertension, diabetes mellitus, and atrial tion findings (left/right hemisphere, cerebellum, brainstem, fibrillation in PCiS has varied widely [2, 7]. The prevalence multiple) as reported by each center’s local radiology depart- of ischemic stroke subtypes in PCiS also varies in previous ment were submitted to the central image repository. reports [1, 2, 8]. An accurate PCiS diagnosis may be difficult to make without appropriate imaging, hence prevalence data MRI evaluation based on clinical diagnosis alone or in combination with computer tomography may be unreliable. Diagnosis of PCiS MRI scans were transferred to the MRI-GENIE neuro- has become easier with the increasing availability of MRI, imaging repository in DICOM format and were available but there are still few large series of MRI-verified PCiS on for review through a secure XNAT viewer [10]. Two sen- which prevalence and etiology data can be based. Few stud- ior neuroradiologists (J.W., M.D.), blinded to the original ies have correlated DWI lesion distribution in the posterior evaluation regarding the presence of DWI lesions, vascular circulation with ischemic stroke subtype. Previous, primarily territory and location, centrally reviewed all MRI images. small, studies have specifically studied infratentorial lesions We excluded cases for poor image quality, lack of DWI or the brainstem alone [9]. Here, we describe a large sample sequences, absence of visible acute ischemic lesion on DWI, of PCiS cases systematically ascertained through radiologi- and the presence of acute ischemic lesions in both vascular cal assessment of the acute lesions on DWI MRI in associa- territories. We grouped patients according to DWI lesion tion with baseline characteristics, vascular risk factors, and location: We defined lesion(s) in the territory of the poste- stroke mechanisms compared to ACiS in the same study. rior cerebral artery (PCA) and its penetrating arteries and in the territories of the vertebral (VA) or basilar arteries (BA) as PCiS. Because we based the stroke phenotypes on DWI Methods lesion location, PCiS was considered present even when a fetal PCA was identified ipsilateral to the acute ischemic Setting lesion. Lesions in the middle cerebral artery (MCA), anterior cerebral artery (ACA), or anterior choroidal artery (ACoA) We reviewed all ischemic stroke cases in the neuroimag- vascular territories were defined as ACiS. The reviewers also ing repository of the MRI-GENetics Interface Exploration recorded the side of lesion (left/right), cerebellar or brain- (MRI-GENIE) collaboration [10]. Data were contributed by stem location in infratentorial lesions, cortical/subcortical 12 of the original National Institute of Neurological Disor- location for supratentorial lesions, and whether lesions were ders and Stroke Genetics Network (NINDS–SiGN) study single or multiple. The reviewers assessed vessel patency sites [11]. Previous publications provide detailed descrip- and evidence of vessel occlusion related to lesion location tions of the NINDS-SiGN and MRI-GENIE studies, as well was assessed in the major intracerebral arteries: the VA, as collection periods and inclusion criteria for each site [10, the BA, the PCA, MCA, and ACA. When neck MRA was 12]. available, the extracranial portion of the VA and the internal carotid arteries (ICA) were assessed regarding the presence Study population of occlusion or stenosis. Reviewers did not specify whether occlusions in the VA were intracranial or extracranial. We Participating sites included five US. sites and seven Euro- did not assess the infratentorial branches of the VA and BA pean sites. Ten sites had hospital-based identification and for presence of occlusion. two sites contributed population-based data. One center con- tributed only data on young stroke patients, aged 15–49, and Stroke mechanisms one center included only patients under the age of 70. Three sites included only first-ever ischemic stroke cases (n = 562) We classified all patients according to the web-based Causa- (Supplementary Table 1 details baseline data for included tive Classification of Stroke system (CCS) [13, 14]. Previous patients per site). publications provide details of the specifics of CCS subtyp- ing within SiGN [12]. Certified physician raters at each GRC Clinical phenotyping adjudicated data on clinical histories and diagnostic testing and fed the information into the web-based CCS system. Each site collected demographic and vascular risk factor data We used five major CCS subtypes for ischemic stroke; car- at the time of enrollment. Clinical phenotype data included dioembolic major (CE), large artery atherosclerosis (LAA), 13 Journal of Neurology (2020) 267:649–658 651 small artery occlusion (SAO), undetermined (UNDETER), Results and other (OTHER). A subset of sites also submitted clas- sification of stroke subtype according to Trial of ORG 10172 At the time of evaluation, the MRI-GENIE database con- in Acute Stroke Treatment (TOAST) [15]. tained MRIs of 3,301 ischemic stroke cases. After exclu- sion for technical or quality reasons, a total of 2,469 cases Analysis with DWI positive lesions on MRI remained. After exclu- sion of patients with acute lesions in both vascular terri- We performed univariable analyses of baseline data for tories (n = 87) or without a decisive allocation of vascular each of the two groups and between groups using the territory (n = 1), 2,381 cases with acute ischemic lesions on Mann–Whitney test for non-parametric variables and the χ2 DWI remained for analysis. The demographic and risk factor test for categorical variables. Missing data points and risk characteristics of the excluded patients with ischemic lesions factor variables recorded as Unknown did not exceed 3% in more than one vascular territory (n = 87) are available for any single variable, except for NIHSS (22%). Missing in Supplementary Table 2. Thirteen patients with posterior data points were not included in any analyses. All vascular lesions and ipsilateral fetal-type PCA were as PCiS. Mag- risk factors, including current smoking, were tested in the netic resonance angiography (MRA) was available for 1,360 univariable analysis. A logistic regression model was cre- patients. Time from stroke onset to scan was recorded in ated for comparison of risk factors between groups. Vari- 89% of cases. A majority (58%) was scanned within 48 h ables that were not significant in the univariable analysis and 86% had been scanned within 14 days. Median time to were removed from the model in the multivariable analysis. scan was 1 day. Statistical analyses were performed using SPSS 23.0 and SPSS 25.0. Baseline characteristics and risk factors Ethics Baseline characteristics, vascular risk factor prevalence and association with respective stroke phenotype for all patients This study and its constituent substudies were conducted are shown in Table 1. in accordance with the Helsinki Declaration as revised in Thirty percent of patients had ischemic lesions in the VB/ 2013. The MRI-GENIE study has been approved by the BA/PCA territories and 70% had lesions in the MCA/ACA institutional review board at Massachusetts’ General Hos- territories on MRI-DWI. Patients with PCiS were younger pital, Boston, MA. All participants provided informed con- than those with ACiS. The proportion of men in the PCiS sent either directly or through surrogate authorization at the group was significantly higher than in ACiS. In men with time of enrollment at the original sites. ischemic stroke, 32% had PCiS vs. 24% of women with ischemic stroke. Male sex was independently associated with Ethical approval PCiS compared with ACiS. Diabetes mellitus was more common in PCiS than in MRI-GENIE IRB number: #: 2001P001186. ACiS and remained independently associated with PCiS Table 1  Vascular risk factor ACiS PCiS Univariable Multivariablea association in ACiS vs. PCiS n = 1 663 (%) n = 718 (%) OR 95% CI OR 95% CI Age (median) 66 63 0.99 0.98–0.99 0.99 0.98–1.00 Male 967 (58) 487 (68) 1.52 1.26–1.82 1.46 1.21–1.78 Hypertensionb 1 096 (66) 451 (64) 0.90 0.75–1.08 – – Diabetes mellitus 373 (23) 190 (27) 1.27 1.03–1.55 1.26 1.02–1.56 Atrial fibrillation 261 (16) 78 (11) 0.64 0.48–0.83 0.70 0.52–0.93 CADb 298 (18) 115 (16) 0.88 0.69–1.11 – – Current smoking 462 (28) 154 (21) 0.71 0.58–0.87 0.63 0.51–0.79 NIHSS (IQR) 4 (2–7) 3 (1–5) – – – – ACiS anterior circulation ischemic stroke, PCiS posterior circulation ischemic stroke, CAD coronary artery disease, NIHSS National Institute of Health Stroke Scale, IQR interquartile range a Logistic regression model adjusting for age, sex, diabetes mellitus, atrial fibrillation, and current smoking b Hypertension and coronary artery disease were not included in the multivariable analysis 13 652 Journal of Neurology (2020) 267:649–658 after adjusting age, sex, atrial fibrillation, and smoking Ischemic stroke subtypes according to CCS status. Atrial fibrillation was less common in patients with PCiS than ACiS. Women with PCiS were significantly Ischemic stroke subtype data according to CCS were avail- more likely to have atrial fibrillation than men (15% vs. 9%, able for all patients. The proportion of patients assigned with p < 0.05). (Supplementary Table 3 details prevalence of risk the CCS subtype UNDETER was similar in both groups factors and CCS in PCiS according to sex). (ACiS 38%, PCiS 39%). The subtype OTHER was assigned in a small number of patients in both groups, with a sig- nificantly higher proportion in PCiS than in ACiS (10% vs. DWI lesion topography in PCiS 6%, p < 0.01). The distribution of specific subtypes (CE, LAA and SAO) differed significantly between ACiS and Among 718 PCiS patients, 197 (27%) had lesions limited PCiS (Fig. 1). The subtypes SAO and LAA were equally to the PCA territory. Cerebellar lesions without concur- common (20%) in PCiS, while LAA was the most common rent brainstem or supratentorial lesions were found in 158 (25%) subtype in ACiS. We studied ischemic stroke sub- (22%) patients. Isolated brainstem lesions without cerebellar type in relation to DWI lesion location in PCiS (Fig. 2). The or supratentorial posterior lesions occurred in 241 patients most common CCS subtype was UNDETER for all locations (33%), and were almost exclusively solitary. Multiple lesions except brainstem location. Single, isolated brainstem lesions limited to the brainstem were rare and occurred in only 28 were attributed to SAO in 47% of the cases and multiple patients (1%). Infratentorial lesions without evidence of lesions limited to the brainstem were attributed to LAA in supratentorial lesions were present in 424 (59%) cases. In 25%. Solitary brainstem lesions were highly predictive for 97 (14%) patients, we found both infratentorial lesions and the SAO subtype (Table 2). lesions in the PCA territory. Large artery atherosclerosis was the second most com- mon stroke subtype in exclusively supratentorial (23%) and isolated cerebellar infarctions (18%) followed by cardioem- Vessel occlusion bolism (15%, respectively). In patients with both cerebellar and/or brainstem and supratentorial lesions, LAA was the MRA sequences were available for 1,360 cases (ACiS, dominant subtype. Among PCiS patients investigated with n = 950; PCiS, n = 410). The proportion of MRA investiga- MRA, the proportion of strokes assigned to the CCS subtype tions did not differ between the groups (57%). Vessel occlu- SAO was significantly lower than in PCiS patients not inves- sion determined to be related to the site of lesion detected tigated with MRA (16% vs. 24%). Conversely, the propor- in 273 (28%) patients with ACiS vs. 159 (39%) with PCiS. tion of patients classified as LAA was higher (23% vs. 16%) Sixty-five (55%) patients among 119 patients with isolated when angiography had been part of the initial assessment. PCA had a visible artery occlusion related to site of lesion. There was no corresponding shift in CCS type assignment The most common sites were the ipsilateral P1–P3 segments. in the ACiS group. Fig. 1  Proportion of ischemic % within vascular territory Anterior Posterior stroke subtypes in ACiS vs. PCiS. ACiS anterior circulation 45 ischemic stroke, PCiS posterior p = 0.662 circulation ischemic stroke, 40 CE cardioembolism, LAA large artery atherosclerosis, SAO 35 small artery occlusion 30 p < 0.01 25 p < 0.001 p < 0.001 20 15 p < 0.01 10 5 0 CE LAA SAO UNDETERMINED OTHER 13 Journal of Neurology (2020) 267:649–658 653 Fig. 2  DWI lesion location % within lesion location CE LAA SAO UNDETERMINED OTHER and ischemic stroke subtype in PCiS. PCiS posterior circulation 50 ischemic stroke, CE cardioem- bolism, LAA large artery ath- 45 erosclerosis, SAO small artery occlusion 40 35 30 25 20 15 10 5 0 Brainstem only Cerebellum only PCA only Cerebellum + PCA n = 241 n = 158 n = 197 n = 55 Table 2  Lesion location and association with CCS subtype SAO (95% CI) LAA (95% CI) CE (95% CI) Undetermined (95% CI) Other (95% CI) Brainstem only OR 9.99 (6.50–15.30)* OR 0.59 (0.30–0.89) OR 0.31 (0.16–0.58)* OR 0.59 (0.43–0.82) OR 0.38 (0.20–0.71) Cerebellum only OR 0.06 (0.02–0.19)* OR 0.90 (0.57–1.41) OR 1.55 (0.93–2.59) OR 1.21 (0.85–1.73) OR 3.99 (2.40–6.63)* PCA only OR 0.62 (0.40–0.97) OR 1.31 (0.88–1.96) OR 1.62 (1.00–2.62) OR 1.17 (0.84–1.63) OR 0.41 (0.21–0.82) Binary logistic regression. CCS subtype vs. all other subtypes for specified anatomical location SAO small artery occlusion, LAA larger artery atherosclerosis, CE cardioembolism *Denotes p < 0.01; regarded as significant to correct for multiple testing Discussion Vascular risk factors In this large sample of MRI-DWI phenotyped cases, we Demographic and risk factor prevalences differed between compared anterior and posterior circulation strokes regard- the two stroke phenotypes. PCiS patients were more often ing risk factor association and ischemic stroke subtypes. male, of younger age, and more frequently diabetic. ACiS We also correlated ischemic stroke subtypes according patients more frequently had atrial fibrillation and were to CCS to lesion location in PCiS. We found that PCiS current smokers. Younger age in PCiS has been reported occurred in 30% of the patients, which is in the upper previously [1, 18–20]. Previous findings regarding hyper- range compared with previously published studies in tension in PCiS vs. ACiS have been divergent: It has been which PCiS prevalence has been reported between 15 reported as a risk factor for PCiS in some studies [20, 21], and 30% depending on stroke subtype classification and while in others, this finding has not been corroborated [1, diagnostic imaging methods [7, 16, 17]. The accuracy 2]. Known risk factor data in PCiS are based on a hetero- of ischemic stroke subtyping and subsequent risk factor geneous set of studies in which the diagnostic certainty assessment has implications in patient management. We varies due to different modes of investigation and diag- believe that this large set of MRI-DWI-verified posterior nostic criteria between studies and often a low number of circulation ischemic stroke cases contributes reliable data PCiS cases. regarding risk factor characteristics and stroke mecha- nisms for this patient group. 13 654 Journal of Neurology (2020) 267:649–658 Specific risk factors in PCiS of the vertebrobasilar tree is performed. This maybe one explanation for the difference in the proportion of patients In multivariable analyses, diabetes mellitus was indepen- classified as SAO in PCiS patients with and without MRA dently associated with PCiS. This is in line with previous in our analysis. Comparisons between this study and other reports of an increased prevalence of PCiS vs. ACiS in dia- registries and studies on PCiS may be unreliable due to the betic patients [1, 19, 22]. Diabetes is a known risk factor for use of different classification systems for stroke subtypes ischemic stroke [23, 24] and is associated with both microan- across studies [16, 21, 33, 34]. Agreement between CCS giopathy and macroangiopathy through complex metabolic and TOAST has been shown to be moderate and differs pathways. Autopsy studies have revealed a higher burden of by subtype [35]. The prevalence of cardioembolic stroke infratentorial ischemic lesions in diabetic patients [25, 26]. in one study [21] was 24%, while 11% of PCiS in our More recently, image-based studies have reported a signifi- study was classified as CE. In keeping with several previ- cant difference in lesion distribution between diabetic and ous studies [7, 22, 36], cardioembolism was significantly non-diabetic patients, with a larger proportion of infarctions more common in ACiS vs. PCiS. The overall prevalence in the VB system in diabetic patients [27, 28]. The reasons of atrial fibrillation in this study (14%) is similar to that in for the higher prevalence of PCiS in diabetic patients remain other large hospital-based ischemic stroke cohorts [37] for unclear with unknown mechanisms for greater susceptibility the same age group. The high (39%) proportion of visible of the vertebrobasilar arteries to diabetes-related injury and vessel occlusions in PCiS on MRA indicates that a large atherogenesis. The prevalence of diabetes mellitus among proportion of PCiS are embolic, especially distal lesions PCiS patients with the ischemic stroke subtypes SAO or affecting the cerebellum and the PCA territory. LAA was 69% vs. 53% in ACiS patients with the same CCS subtypes. This may indicate that there are mechanisms of diabetic injury specifically affecting the vertebrobasilar Limitations arteries in ways that remain to be elucidated. A recent study aiming to establish causative links between type 2 diabetes This study has several limitations. Not all patients pre- (T2D) and specific ischemic stroke subtypes showed a firm senting with clinical stroke and subsequently enrolled at association for T2D with large artery stroke, but not with each study center had MRI imaging. The inclusion in our small artery stroke [29]. study of only cases with DWI lesions may have led to The complex influence of biological sex on ischemic selection bias in terms of stroke severity, if patients with stroke risk was beyond the scope of our current study. How- very severe stroke were not considered for MRI investiga- ever, male sex as risk factor for PCiS vs. ACiS has been tions. It may also have resulted in the exclusion of genuine reported previously [19–21]. A male preponderance in ischemic stroke cases, since negative DWI in the acute PCiS has been observed in both children [30] and young stage occurs, particularly with smaller lesions and may dif- adults [5]. There are studies indicating sex differences in the fer between anterior and posterior locations [38]. Negative autoregulatory capacity of the basilar artery in young teens DWI is more common in PCiS than ACiS, in part due to [31] and a significantly lower cerebrovascular reactivity to the smaller volume of perfused brain parenchyma in the l-arginine in the posterior circulation in adult males [32]. posterior fossa, but also because image artifacts are more common in this location than in the anterior circulation Lesion location and CCS subtype [39]. A potential limitation is the exclusion of 87 patients We explored the lesion location and CCS subtype sepa- with acute ischemic lesions in more than one vascular ter- rately for the PCA territory, the cerebellum and the brain- ritory. Since cardiac embolism as a causative mechanism stem. In isolated brainstem lesions, the SAO subtype pre- is often overrepresented in patients with acute lesions dominated. Lesions in the PCA territory and cerebellum in multiple vascular territories, exclusion of this group were most often related to LAA or CE. This is in keep- (n = 87) from our analyses could potentially have led to ing with previous reports [16, 33, 34]. In our study, SAO a false high proportion of other causative mechanisms and LAA were equally common in PCiS overall, but on or risk factors, such as diabetes, in the PCiS group. To analysis of the relationship between location and ischemic explore this potential bias, we performed a sensitivity stroke subtype, we found that exclusive, solitary brainstem analysis in which the 87 patients intended for exclusion lesions are heavily associated with and highly predictive of were added to the PCiS group and compared to the ACiS small artery occlusion. However, small brainstem lesions group using the same statistical methods as for the main may be caused by occlusions in branch arteries, or severe analysis. The results of the sensitivity analysis did not stenosis of the basilar artery and may be routinely misclas- differ significantly from the results of the main analyses sified as small artery occlusion, unless vascular imaging presented in Table 1. 13 Journal of Neurology (2020) 267:649–658 655 Strengths concept and design, drafting the manuscript, and critical revision for intellectual content. The strict stroke phenotyping and definitive PCiS diagno- Funding Dr. Roquer: supported in part by Spain’s Ministry of Health sis achieved by verification of DWI positive lesions, and (Ministerio de Sanidad y Consumo, Instituto de Salud Carlos III the high number of included cases are the major strengths FEDER, RD12/0042/0020). Dr. Jern: The Swedish Research Council of our study. Enrollment at the individual study sites was (2018-02543), The Swedish Heart and Lung Foundation (20160316) as well as grants from the Swedish state under the agreement between based on all ischemic stroke, regardless of vascular terri- the Swedish government and the county councils, the ALF-agreement tory involved, and therefore the comparison between ACiS (ALFGBG-720081). Dr. Meschia: receives funding for the CREST-2 and PCiS in our study may be generalizable to ischemic Trial from the NINDS U01 NS080168 and from Mayo Clinic through stroke populations in secondary and tertiary stroke centers. the Earl and Nyda Swanson Neurosciences Research Fund and the Harley N. and Rebecca N. Hotchkiss Endowed Fund in Neuroscience Research, Honoring Ken and Marietta. Dr. Kittner: supported in part by award number U01NS069208 from the National Institute of Neurologi- cal Disorders and Stroke (NINDS). The content is solely the respon- Conclusions sibility of the authors and does not necessarily represent the official views of the NINDS or the National Institutes of Health. Dr. Kittner’s effort on this project was partly supported by VA RR&D N1699-R and PCiS is a heterogeneous stroke phenotype with distinct NINDS R01NS100178 and R01NS105150. Dr. Mitchell: supported features related to vascular supply, lesion location, and in part by SiGN grant for funding R01NS100178. Dr. Rosand: sup- associated ischemic stroke subtype. The proportion of ves- ported in part by US NIH-NINDS R01NS100178 and R01NS086905. Dr. Golland: NIH NIBIB NAC P41EB015902. Dr. Rost: NIH-NINDS sel occlusion observed in PCiS is high. Visualization of R01 NS086905 (MRI-GENIE study/Rost PI). Dr. Lindgren: the Lund the VB vascular tree and MRI lesion site(s) and distribu- Stroke Register has been supported by the Swedish Heart and Lung tion may aid in determining likely clinical stroke subtype Foundation, Skåne University Hospital, Region Skåne, the Freemasons in PCiS, which may influence selection of treatment of Lodge of Instruction EOS in Lund, Foundation of Färs and Frosta (one of Sparbanken Skåne’s ownership Foundations), and Lund University. risk factors and secondary prevention strategies. Our study indicates an increased risk of PCiS vs. ACiS in diabetic patients and in males after adjusting for other conventional Compliance with ethical standards stroke risk factors. Genetically determined molecular path- Conflicts of interest PF, MD, AKG, JW, MDS, KLD, LC, RI, MJRJB, ways may specifically influence pathogenic processes in ECM, AVD, RS, HX, EGS, LH, KJ, J Roquer, JWC, PFMA, JPB, JJC, the vertebrobasilar arteries. MRI-based phenotyping of CJ, BMK, DOK, RL, JFM, TR, RLS, RS, PS, AS, DW, BBW, SJK, PCiS vs. ACiS may contribute valuable insights into spe- BDM, PG, and JP: None. SJTM reports personal fees from Alphabet, Inc., outside the submitted work; VT reports grants from NINDS fund- cific risk factor profiles and mechanisms of injury in PCiS. ing, during the conduct of the study; personal fees and non-financial support from Boehringer Ingelheim, personal fees from Pfizer, per- Acknowledgements Open access funding provided by Lund Uni- sonal fees from Medtronic, personal fees from Amgen, personal fees versity. The authors would like to extend their gratitude to the many from Bayer, outside the submitted work; J Rosand reports personal fees patients who consented to the use of their clinical and image data. We from Boehringer Ingelheim, Pfizer, and New Beta Innovations, outside would like to thank Peter Höglund, MD, PhD for statistical advice. the submitted work; OW reports grants from NIH-NINDS, during the MDS would like to acknowledge funding from the European Union’s conduct of the study; personal fees from Penumbra, Inc and Genen- Horizon 2020 research and innovation program under the Marie Sklo- tech, outside the submitted work; NSR reports grants from National dowska-Curie Grant agreement no 753896, outside the submitted work. Institutes of Health, during the conduct of the study; AL reports per- sonal fees from Bayer, BMS/Pfizer, Astra Zeneca, and Boehringer In- Author contributions PF: study concept and design, data base design, gelheim, outside the submitted work. data analysis, drafting, and revising the manuscript; AKG: data analy- sis, critical revision for intellectual content; MD, JW: review and analy- Open Access This article is distributed under the terms of the Crea- sis of MRI data, data base management, critical revision for intellectual tive Commons Attribution 4.0 International License (http://creat​iveco​ content; MDS: maintenance of imaging data, critical revision for intel- mmons​.org/licen​ses/by/4.0/), which permits unrestricted use, distribu- lectual content; KLD and LC: maintenance of imaging data and critical tion, and reproduction in any medium, provided you give appropriate revision for intellectual content; RI, MJRJB, ECM, SJTM, AVD, and credit to the original author(s) and the source, provide a link to the RS: maintenance of imaging database and critical revision for intellec- Creative Commons license, and indicate if changes were made. tual content; HX: acquisition of data and critical revision for intellec- tual content; EGS, JRoq: acquisition of imaging data and critical revi- sion for intellectual content; LH, KJ, JWC, PFM, JPB, JJC, CJ, BMK, DOK, RL, JFM, TR, RLS, RS, PS, AS, VT, DW, and BBW: acquisition References of data and critical revision for intellectual content; SJK, BDM, JRos, and JP; study concept and critical revision for intellectual content; PG: 1. Subramanian G, Silva J, Silver FL, Fang J, Kapral MK, Oczkowski acquisition of data and critical revision for intellectual content; OW: W, Gould L, O’Donnell MJ, Investigators of the Registry of the acquisition of imaging data, review and analysis of MRI data, critical Canadian Stroke N (2009) Risk factors for posterior compared to revision for intellectual content; NSR: study concept and design, and anterior ischemic stroke: an observational study of the Registry of critical revision for intellectual content; AL: study supervision, study 13 656 Journal of Neurology (2020) 267:649–658 the Canadian Stroke Network. Neuroepidemiology 33(1):12–16. study of ischemic stroke subtypes. Stroke 44(10):2694–2702. https​://doi.org/10.1159/00020​9282 https​://doi.org/10.1161/strok​eaha.113.00185​7 2. De Marchis GM, Kohler A, Renz N, Arnold M, Mono ML, Jung 13. Ay H, Benner T, Arsava EM, Furie KL, Singhal AB, Jensen S, Fischer U, Karameshev AI, Brekenfeld C, Gralla J, Schroth G, MB, Ayata C, Towfighi A, Smith EE, Chong JY, Koroshetz WJ, Mattle HP, Nedeltchev K (2011) Posterior versus anterior circu- Sorensen AG (2007) A computerized algorithm for etiologic lation strokes: comparison of clinical, radiological and outcome classification of ischemic stroke: the Causative Classification of characteristics. J Neurol Neurosurg Psychiatry 82(1):33–37. https​ Stroke System. Stroke 38(11):2979–2984. https:​ //doi.org/10.1161/ ://doi.org/10.1136/jnnp.2010.21115​1 STROK​EAHA.107.49089​6 3. Bogousslavsky J, Van Melle G, Regli F (1988) The Lausanne 14. Arsava EM, Ballabio E, Benner T, Cole JW, Delgado-Martinez Stroke Registry: analysis of 1,000 consecutive patients with first MP, Dichgans M, Fazekas F, Furie KL, Illoh K, Jood K, Kittner stroke. Stroke 19(9):1083–1092 S, Lindgren AG, Majersik JJ, Macleod MJ, Meurer WJ, Mon- 4. Caplan LR, Wityk RJ, Glass TA, Tapia J, Pazdera L, Chang HM, taner J, Olugbodi AA, Pasdar A, Redfors P, Schmidt R, Sharma Teal P, Dashe JF, Chaves CJ, Breen JC, Vemmos K, Amarenco P, P, Singhal AB, Sorensen AG, Sudlow C, Thijs V, Worrall BB, Tettenborn B, Leary M, Estol C, Dewitt LD, Pessin MS (2004) Rosand J, Ay H, International Stroke Genetics C (2010) The New England Medical Center Posterior Circulation Registry. Ann Causative Classification of Stroke system: an international reli- Neurol 56(3):389–398. https​://doi.org/10.1002/ana.20204​ ability and optimization study. Neurology 75(14):1277–1284. 5. von Sarnowski B, Schminke U, Grittner U, Tanislav C, Bottcher https​://doi.org/10.1212/wnl.0b013​e3181​f612c​e T, Hennerici MG, Tatlisumak T, Putaala J, Kaps M, Fazekas F, 15. Adams HP Jr, Bendixen BH, Kappelle LJ, Biller J, Love BB, Enzinger C, Rolfs A, Kessler C, Sifap I (2017) Posterior versus Gordon DL, Marsh EE 3rd (1993) Classification of subtype of anterior circulation stroke in young adults: a comparative study acute ischemic stroke Definitions for use in a multicenter clini- of stroke aetiologies and risk factors in stroke among young fabry cal trial. TOAST. Trial of Org 10172 in Acute Stroke Treatment. patients (sifap1). Cerebrovasc Dis 43(3–4):152–160. https​://doi. Stroke 24(1):35–41 org/10.1159/00045​4840 16. Moulin T, Tatu L, Vuillier F, Berger E, Chavot D, Rumbach L 6. Markus HS, van der Worp HB, Rothwell PM (2013) Posterior (2000) Role of a stroke data bank in evaluating cerebral infarc- circulation ischaemic stroke and transient ischaemic attack: diag- tion subtypes: patterns and outcome of 1,776 consecutive nosis, investigation, and secondary prevention. Lancet Neurol patients from the Besancon stroke registry. Cerebrovasc Dis 12(10):989–998. https:​ //doi.org/10.1016/S1474-​ 4422(13)70211-​ 4 10(4):261–271. https​://doi.org/10.1159/00001​6068 7. Di Carlo A, Lamassa M, Baldereschi M, Pracucci G, Consoli 17. Cates MJ, Paton JF, Smeeton NC, Wolfe CD (2012) Hyperten- D, Wolfe CD, Giroud M, Rudd A, Burger I, Ghetti A, Inzitari sion before and after posterior circulation infarction: analysis of D, European BSoSCG (2006) Risk factors and outcome of sub- data from the South London Stroke Register. J Stroke Cerebro- types of ischemic stroke. Data from a multicenter multinational vasc Dis 21(7):612–618. https​: //doi.org/10.1016/j.jstro​kecer​ hospital-based registry. The European Community Stroke Pro- ebrov​asdis​.2011.02.004 ject. J Neurol Sci 244(1–2):143–150. https​://doi.org/10.1016/j. 18. Dornak T, Kral M, Hazlinger M, Herzig R, Veverka T, Burval jns.2006.01.016 S, Sanak D, Zapletalova J, Antalikova K, Kanovsky P (2015) 8. Libman RB, Kwiatkowski TG, Hansen MD, Clarke WR, Woolson Posterior vs. anterior circulation infarction: demography, out- RF, Adams HP (2001) Differences between anterior and posterior comes, and frequency of hemorrhage after thrombolysis. Int J circulation stroke in TOAST. Cerebrovasc Dis 11(4):311–316. Stroke 10(8):1224–1228. https​://doi.org/10.1111/ijs.12626​ https​://doi.org/10.1159/00004​7659 19. Tao WD, Liu M, Fisher M, Wang DR, Li J, Furie KL, Hao 9. Hong YH, Zhou LX, Yao M, Zhu YC, Cui LY, Ni J, Peng B (2018) ZL, Lin S, Zhang CF, Zeng QT, Wu B (2012) Posterior versus Lesion topography and its correlation with etiology in medullary anterior circulation infarction: how different are the neurologi- infarction: analysis from a multi-center stroke study in China. cal deficits? Stroke 43(8):2060–2065. https​://doi.org/10.1161/ Front Neurol 9:813. https​://doi.org/10.3389/fneur​.2018.00813​ STROK​EAHA.112.65242​0 10. Giese AK, Schirmer MD, Donahue KL, Cloonan L, Irie R, 20. Miyamoto N, Tanaka Y, Ueno Y, Tanaka R, Hattori N, Urabe Winzeck S, Bouts M, McIntosh EC, Mocking SJ, Dalca AV, T (2010) Comparison of clinical backgrounds with anterior Sridharan R, Xu H, Frid P, Giralt-Steinhauer E, Holmegaard versus posterior circulation infarcts. J Stroke Cerebrovasc Dis L, Roquer J, Wasselius J, Cole JW, McArdle PF, Broderick JP, 19(5):393–397. https​://doi.org/10.1016/j.jstro​kecer​ebrov​asdis​ Jimenez-Conde J, Jern C, Kissela BM, Kleindorfer DO, Lem- .2009.07.012 mens R, Lindgren A, Meschia JF, Rundek T, Sacco RL, Schmidt 21. Caplan L, Chung CS, Wityk R, Glass T, Tapia J, Pazdera L, Chang R, Sharma P, Slowik A, Thijs V, Woo D, Worrall BB, Kittner SJ, HM, Dashe J, Chaves C, Vemmos K, Leary M, Dewitt L, Pessin M Mitchell BD, Rosand J, Golland P, Wu O, Rost NS (2017) Design (2005) New England Medical Center Posterior Circulation Stroke and rationale for examining neuroimaging genetics in ischemic Registry: I. Methods, data base, distribution of brain lesions, stroke: the MRI-GENIE study. Neurol Genet 3(5):e180. https​:// stroke mechanisms, and outcomes. J Clin Neurol 1(1):14–30. https​ doi.org/10.1212/NXG.00000​00000​00018​0 ://doi.org/10.3988/jcn.2005.1.1.14 11. Network NSG, International Stroke Genetics C (2016) Loci asso- 22. Zeng Q, Tao W, Lei C, Dong W, Liu M (2015) Etiology and risk ciated with ischaemic stroke and its subtypes (SiGN): a genome- factors of posterior circulation infarction compared with anterior wide association study. Lancet Neurol 15(2):174–184. https​://doi. circulation infarction. J Stroke Cerebrovasc Dis 24(7):1614–1620. org/10.1016/S1474​-4422(15)00338​-5 https​://doi.org/10.1016/j.jstro​kecer​ebrov​asdis​.2015.03.033 12. Meschia JF, Arnett DK, Ay H, Brown RD Jr, Benavente OR, Cole 23. Kannel WB, McGee DL (1979) Diabetes and cardiovascular dis- JW, de Bakker PI, Dichgans M, Doheny KF, Fornage M, Grewal ease. The Framingham study. JAMA 241(19):2035–2038 RP, Gwinn K, Jern C, Conde JJ, Johnson JA, Jood K, Laurie CC, 24. Emerging Risk Factors C, Sarwar N, Gao P, Seshasai SR, Gobin Lee JM, Lindgren A, Markus HS, McArdle PF, McClure LA, R, Kaptoge S, Di Angelantonio E, Ingelsson E, Lawlor DA, Selvin Mitchell BD, Schmidt R, Rexrode KM, Rich SS, Rosand J, Roth- E, Stampfer M, Stehouwer CD, Lewington S, Pennells L, Thomp- well PM, Rundek T, Sacco RL, Sharma P, Shuldiner AR, Slowik son A, Sattar N, White IR, Ray KK, Danesh J (2010) Diabetes A, Wassertheil-Smoller S, Sudlow C, Thijs VN, Woo D, Worrall mellitus, fasting blood glucose concentration, and risk of vascular BB, Wu O, Kittner SJ, Study NS (2013) Stroke Genetics Network disease: a collaborative meta-analysis of 102 prospective studies. (SiGN) study: design and rationale for a genome-wide association 13 Journal of Neurology (2020) 267:649–658 657 Lancet 375(9733):2215–2222. https​://doi.org/10.1016/S0140​ 33. Yamamoto Y, Georgiadis AL, Chang HM, Caplan LR (1999) -6736(10)60484​-9 Posterior cerebral artery territory infarcts in the New England 25. Aronson SM (1973) Intracranial vascular lesions in patients with Medical Center Posterior Circulation Registry. Arch Neurol diabetes mellitus. J Neuropathol Exp Neurol 32(2):183–196 56(7):824–832 26. Peress NS, Kane WC, Aronson SM (1973) Central nervous system 34. Vemmos KN, Takis CE, Georgilis K, Zakopoulos NA, Lekakis JP, findings in a tenth decade autopsy population. Prog Brain Res Papamichael CM, Zis VP, Stamatelopoulos S (2000) The Athens 40:473–483. https​://doi.org/10.1016/S0079​-6123(08)60707​-4 stroke registry: results of a five-year hospital-based study. Cer- 27. Iwase M, Yamamoto M, Yoshinari M, Ibayashi S, Fujishima M ebrovasc Dis 10(2):133–141. https​://doi.org/10.1159/00001​6042 (1998) Stroke topography in diabetic and nondiabetic patients 35. McArdle PF, Kittner SJ, Ay H, Brown RD Jr, Meschia JF, Rundek by magnetic resonance imaging. Diabetes Res Clin Pract T, Wassertheil-Smoller S, Woo D, Andsberg G, Biffi A, Brenner 42(2):109–116 DA, Cole JW, Corriveau R, de Bakker PI, Delavaran H, Dichgans 28. Arboix A, Rivas A, Garcia-Eroles L, de Marcos L, Massons J, M, Grewal RP, Gwinn K, Huq M, Jern C, Jimenez-Conde J, Jood Oliveres M (2005) Cerebral infarction in diabetes: clinical pat- K, Kaplan RC, Katschnig P, Katsnelson M, Labovitz DL, Lem- tern, stroke subtypes, and predictors of in-hospital mortality. BMC mens R, Li L, Lindgren A, Markus HS, Peddareddygari LR, Ped- Neurol 5(1):9. https​://doi.org/10.1186/1471-2377-5-9 ersen A, Pera J, Redfors P, Roquer J, Rosand J, Rost NS, Rothwell 29. Larsson SC, Scott RA, Traylor M, Langenberg CC, Hindy G, Mel- PM, Sacco RL, Sharma P, Slowik A, Sudlow C, Thijs V, Tiedt ander O, Orho-Melander M, Seshadri S, Wareham NJ, Markus S, Valenti R, Worrall BB, Study NS (2014) Agreement between HS, Collaboration M, Network NSG (2017) Type 2 diabetes, TOAST and CCS ischemic stroke classification: the NINDS SiGN glucose, insulin, BMI, and ischemic stroke subtypes: mende- study. Neurology 83(18):1653–1660. https​://doi.org/10.1212/ lian randomization study. Neurology 89(5):454–460. https​://doi. wnl.00000​00000​00094​2 org/10.1212/WNL.00000​00000​00417​3 36. Zurcher E, Richoz B, Faouzi M, Michel P (2018) Differences 30. Ganesan V, Chong WK, Cox TC, Chawda SJ, Prengler M, in ischemic anterior and posterior circulation strokes: a clinico- Kirkham FJ (2002) Posterior circulation stroke in childhood: risk radiological and outcome analysis. J Stroke Cerebrovasc Dis. https​ factors and recurrence. Neurology 59(10):1552–1556 ://doi.org/10.1016/j.jstro​kecer​ebrov​asdis​.2018.11.016 31. Vavilala MS, Kincaid MS, Muangman SL, Suz P, Rozet I, Lam 37. RIKSSTROKE The Swedish Stroke Register. http://www.rikss​ AM (2005) Gender differences in cerebral blood flow velocity troke​.org/. Accessed Sept 2018 and autoregulation between the anterior and posterior circula- 38. Oppenheim C, Stanescu R, Dormont D, Crozier S, Marro B, Sam- tions in healthy children. Pediatr Res 58(3):574–578. https​://doi. son Y, Rancurel G, Marsault C (2000) False-negative diffusion- org/10.1203/01.PDR.00001​79405​.30737​.0F weighted MR findings in acute ischemic stroke. AJNR Am J Neu- 32. Perko D, Pretnar-Oblak J, Sabovic M, Zvan B, Zaletel M (2011) roradiol 21(8):1434–1440 Differences between cerebrovascular reactivity to l-arginine in 39. Muir KW, Buchan A, von Kummer R, Rother J, Baron JC (2006) the anterior and posterior cerebral circulation. Cerebrovasc Dis Imaging of acute stroke. Lancet Neurol 5(9):755–768. https​://doi. 31(4):358–364. https​://doi.org/10.1159/00032​2562 org/10.1016/S1474​-4422(06)70545​-2 Affiliations Petrea Frid1,2,3 · Mattias Drake4,5 · A. K. Giese6,7 · J. Wasselius4,5 · M. D. Schirmer6,8,9 · K. L. Donahue6 · L. Cloonan6 · R. Irie10 · M. J. R. J. Bouts10 · E. C. McIntosh10 · S. J. T. Mocking10 · A. V. Dalca8,10 · R. Sridharan8 · H. Xu11 · E. Giralt‑Steinhauer12 · L. Holmegaard13,14 · K. Jood13,14 · J. Roquer12 · J. W. Cole15 · P. F. McArdle11 · J. P. Broderick16 · J. Jimenez‑Conde12 · C. Jern17 · B. M. Kissela16 · D. O. Kleindorfer16 · R. Lemmens18,19,20 · J. F. Meschia21 · T. Rundek22 · R. L. Sacco22 · R. Schmidt23 · P. Sharma24,25 · A. Slowik26 · V. Thijs27,28 · D. Woo16 · B. B. Worrall29,30 · S. J. Kittner15 · B. D. Mitchell11,31 · J. Petersson1,2 · J. Rosand6,7,10,32,33 · P. Golland8 · O. Wu10 · N. S. Rost6 · A. Lindgren1,34 on behalf of the Stroke Genetics Network (SiGN), the International Stroke Genetics Consortium (ISGC), and the MRI-Genetics Interface Exploration (MRI-GENIE) Study 1 7 Department of Clinical Sciences Lund, Neurology, Lund Program in Medical and Population Genetics, Broad Institute University, Lund, Sweden of MIT and Harvard, Cambridge, MA, USA 2 8 Department of Neurology and Rehabilitation Medicine, Computer Science and Artificial Intelligence Laboratory, Neurology, Skåne University Hospital, Malmö, Sweden MIT, Cambridge, USA 3 9 Department of Neurology, Skåne University Hospital, Jan Department of Population Health Sciences, German Centre Waldenströms gata 19, 205 02 Malmö, Sweden for Neurodegenerative Diseases (DZNE), Bonn, Germany 4 10 Department of Clinical Sciences Lund, Radiology, Lund Department of Radiology, Athinoula A. Martinos Center University, Lund, Sweden for Biomedical Imaging, Massachusetts General Hospital 5 (MGH), Harvard Medical School, Charlestown, MA, USA Department of Radiology, Neuroradiology, Skåne University 11 Hospital, Lund, Sweden Division of Endocrinology, Diabetes and Nutrition, 6 Department of Medicine, University of Maryland School J. Philip Kistler Stroke Research Center, Department of Medicine, Baltimore, MD, USA of Neurology, Massachusetts General Hospital, Harvard 12 Medical School, Boston, MA, USA Neurovascular Research Group (NEUVAS), Department of Neurology, IMIM‑Hospital del Mar (Institut Hospital del 13 658 Journal of Neurology (2020) 267:649–658 23 Mar d’Investigacions Mèdiques), Universitat Autonoma de Clinical Division of Neurogeriatrics, Department Barcelona, Barcelona, Spain of Neurology, Medical University Graz, Graz, Austria 13 24 Department of Clinical Neuroscience, Institute Institute of Cardiovascular Research, Royal Holloway of Neuroscience and Physiology, The Sahlgrenska Academy University of London (ICR2UL), Egham, UK at University of Gothenburg, Gothenburg, Sweden 25 Ashford and St Peter’s Hospital, Ashford, UK 14 Department of Neurology, Sahlgrenska University Hospital, 26 Department of Neurology, Jagiellonian University Medical Gothenburg, Sweden College, Kraków, Poland 15 Department of Neurology, University of Maryland School 27 Stroke Division, Florey Institute of Neuroscience and Mental of Medicine and Veterans Affairs Maryland Health Care Health, Heidelberg, Australia System, Baltimore, MD, USA 28 16 Department of Neurology, Austin Health, Heidelberg, Department of Neurology and Rehabilitation Medicine, Australia University of Cincinnati College of Medicine, Cincinnati, 29 OH, USA Department of Neurology, University of Virginia, 17 Charlottesville, VA, USA Institute of Biomedicine, The Sahlgrenska Academy 30 at University of Gothenburg, Gothenburg, Sweden Department of Public Health Sciences, University 18 of Virginia, Charlottesville, VA, USA Department of Neurosciences, Experimental Neurology, KU 31 Leuven-University of Leuven, Louvain, Belgium Geriatric Research and Education Clinical Center, Veterans 19 Administration Medical Center, Baltimore, MD, USA VIB Center for Brain and Disease Research, Louvain, 32 Belgium Center for Genomic Research, Massachusetts General 20 Hospital, Boston, MA, USA Department of Neurology, University Hospitals Leuven, 33 Louvain, Belgium Division of Neurocritical Care and Emergency Neurology, 21 Massachusetts General Hospital, Boston, MA, USA Department of Neurology, Mayo Clinic, Jacksonville, FL, 34 USA Department of Neurology and Rehabilitation Medicine, 22 Neurology, Skåne University Hospital, Lund, Sweden Department of Neurology and Evelyn F. McKnight Brain Institute, Miller School of Medicine, University of Miami, Miami, FL, USA 13