Abstract
Background
Intracranial hemorrhage (ICH) secondary to hypertension (HTN) classically occurs in the basal ganglia, cerebellum, or pons. Vascular lesions such as aneurysms or arteriovenous malformations (AVMs) are more common in younger patients. We investigated the utility of diagnostic subtraction angiography (DSA) in young hypertensive patients with non-lobar ICH.
Methods
A retrospective review (2013–2022) identified young (18–60 years) patients who underwent DSA for ICH. HTN history, ICH location, presence/absence of subarachnoid hemorrhage (SAH), and computed tomography angiography (CTA) findings were collected. The main outcome was DSA-positivity, defined as presence of an AVM, aneurysm, Moyamoya disease, reversible cerebral vasoconstriction syndrome, or dural arteriovenous fistula on DSA.
Results
Two hundred sixty patients were included, and the DSA-positivity rate was 19%.
DSA-positivity was lower in hypertensive patients with ICHs in the cerebellum, pons, or basal ganglia compared to the rest of the patient sample (9% vs 26%, p = 0.0002, Fisher's exact test). We developed the ICH-Angio score (0–5 points) based on CTA findings, ICH location, HTN history, and presence of SAH to predict risk of underlying vascular lesions. DSA-positivity was lower in those with a score of 0 (0/62; 0%) compared to a score of 1 (5/52; 10%), 2 (17/48; 35%), 3 (10/20; 50%), 4 (5/6; 83%), or 5 (3/3; 100%).
Conclusion
The ICH-Angio score was able to non-invasively rule out an underlying vascular etiology for ICH in up to one-third of patients. HTN, ICH location, CTA findings, and associated SAH can identify patients at low risk for harboring underlying vascular lesions.
Keywords: Angiography, intracranial hemorrhage, diagnostic accuracy, vascular lesion, digital subtraction angiography
Introduction
Intracranial hemorrhage (ICH) is a devastating neurological disorder that carries significant morbidity and mortality and impacts approximately 40,000 to 67,000 Americans every year. 1 Etiologies of ICH are numerous and include arteriovenous malformation (AVM), aneurysms, intracranial tumors, amyloid angiopathy, anticoagulation, Moyamoya disease, and trauma. 2 Diagnostic imaging for ICH includes computed tomography angiography (CTA), magnetic resonance imaging, and cerebral angiography. Although diagnostic cerebral angiography is the gold standard for diagnosis of vascular lesions, it is an invasive modality that carries non-negligible periprocedural risk such as acute kidney injury (AKI), transient ischemic attack and stroke, which patients with hypertension (HTN) may be uniquely susceptible to.3–5
Currently, it is unclear which patients will benefit the most from diagnostic cerebral angiography for spontaneous ICH. One study by Zhu et al. demonstrated that the rate of underlying vascular etiologies for non-lobar ICH was 0% in both young ( ≤ 45 years old) and old (>45 years) patients with HTN. 5 This finding suggests that the diagnostic yield of diagnostic subtraction angiography (DSA) is minimal in patients with HTN and spontaneous ICH in locations classically associated with HTN, regardless of age. Another study performed a systematic review of 12 studies, which suggested that ICH location and HTN history, not age, were the most predictive factors of diagnosing underlying vascular etiologies. However, that same study surveyed 692 physicians and showed that age was the single most important factor in clinician determination for which patients to select for DSA. These prior studies suggest that there are unique clinical and radiographic features that may be able to discriminate between young patients with and without underlying vascular causes of ICH. To that end, the goal of this study is to identify predictors of positive angiography findings of an underlying vascular lesion specifically in young patients (≤ 60 years old) with ICH. We hypothesized that cerebral angiography would have limited diagnostic value for patients who presented with non-lobar ICH with negative pre-DSA CTA and a lack of subarachnoid hemorrhage (SAH).
Methods
Patient selection
This study was approved by our institutional review board, conforms to the STROBE statement, 6 and did not require patient consent due to its retrospective nature. All cases of spontaneous ICH admitted to the neurosurgical service of our institution from 2013 to 2022 were reviewed. Exclusion criteria included patients > 60 and < 18 years of age, those without intraparenchymal hemorrhage (e.g. primary intraventricular hemorrhage), traumatic ICHs, and those who did not undergo a diagnostic cerebral angiogram. The primary outcome was DSA-positivity, defined as patients who were diagnosed with AVM, aneurysm, Moyamoya disease, reversible cerebral vasoconstriction syndrome (RCVS), or dural arteriovenous fistula (dAVF) on DSA. These disease processes were identified as the main outcome as they were potentially treatable with medication, endovascular, or surgical means.
Data collection
Collected variables included patient baseline characteristics such as age, gender, race, smoking history, usage of anticoagulant/antiplatelet medication, periprocedural complications associated with cerebral angiography, and presence of HTN. Notably, HTN was defined as having a systolic blood pressure > 160 mmHg on presentation or having a documented past medical history of HTN requiring antihypertensive medications. Collected radiographic data included ICH location (cortex, basal ganglia, pons, or cerebellum), CTA, if performed, presence or absence of SAH, and diagnostic cerebral angiogram findings. CTA findings were characterized as either negative, positive, or suspicious for an underlying vascular etiology. A suspicious CTA was identified if there were findings suggestive of an underlying vascular cause of ICH but could not be definitively visualized.
Statistical analysis
We identified a total of 260 patients for statistical analysis. Categorical variables were compared using Chi-squared or Fisher exact test. Continuous variables were compared using a Student's t-test or a Mann–Whitney Wilcoxon Test depending on the normality of the data. A novel scoring system (the ICH-Angio score) that incorporated CTA findings (negative, suspicious, or positive for vascular lesion) [0–2 points], ICH location (basal ganglia/pons/cerebellum versus cortical) [0–1 points], HTN history (yes versus no) [0–1 points], and SAH (presence versus absence) [0–1 points] was created to predict a given ICH patient's probability of having an underlying vascular lesion. A linear-by-linear association test was used to assess the correlation between risk score and diagnostic cerebral angiography positivity rate. Of note, when calculating the sensitivity and specificity of CTA to detect an underlying vascular lesion, a suspicious CTA finding was considered equivalent to a positive CTA finding. All statistical analyses were performed using JMP 16 (SAS, Singapore) and a p-value ≤ 0.05 was deemed significant unless otherwise indicated.
Results
A total of 260 patients with acute ICH were analyzed. The mean age in years (interquartile range) was 46 (37–56) and a slight majority of the cohort were males (55%) and Caucasians (57%). The most common ICH locations included the cortex (n = 130; 50%), basal ganglia (n = 94; 36%), cerebellum (n = 25; 10%), and pons (n = 7; 3%). The majority of patients (n = 194; 75%) had HTN and a summary of the baseline demographics are included in Table 1. After assessing all angiograms, 50 patients (19%) had evidence of an underlying vascular lesion on DSA. The underlying vascular lesions consisted of AVMs (n = 22), aneurysms (n = 17), RCVS (n = 6), Moyamoya disease (n = 4), and dAVF (n = 1) (Table 2). Eleven patients (4%) had DSA-related complications, which all consisted of contrast induced AKI.
Table 1.
Baseline demographics.
| Variable a | N = 260 |
|---|---|
| Age | 45.88 (37–56) |
| Gender (male, %) | 143 (55%) |
| Race | |
| Caucasian | 149 (57%) |
| African American | 91 (35%) |
| Other | 20 (8%) |
| ICH location | |
| Basal ganglia | 94 (36%) |
| Cortical | 130 (50%) |
| Cerebellum | 25 (10%) |
| Pontine | 7 (3%) |
| Other | 4 (2%) |
| Hypertensive (n, %) | 194 (75%) |
Reported as mean (interquartile range) for continuous variables.
Table 2.
Final diagnoses for patients with a positive DSA.
| Diagnosis | N = 50 |
|---|---|
| AVM | 22 (44%) |
| Aneurysm | 17 (34%) |
| Moya-Moya disease | 4 (8%) |
| Fistula | 1 (2%) |
| RCVS | 6 (12%) |
AVM: arteriovenous malformation; RCVS: reversible cerebral vasoconstriction syndrome.
On univariate analysis, there was no significant difference between those with an underlying vascular lesion on diagnostic cerebral angiography and those without a lesion with respect to age (p = 0.36), gender (p = 0.30), race (0.42), smoking status (p = 0.44), and rates of CTA acquisition prior to cerebral angiography (p = 0.21). However, those without an underlying vascular lesion were more likely to be hypertensive on presentation or have a previous diagnosis of HTN (79% versus 56%, p = 0.0013). Furthermore, patients with an underlying vascular lesion were more likely to present with a cortical ICH (68% versus 46%) and less likely to present with a basal ganglia ICH (18% versus 40%) (p = 0.0081). SAH was more prevalent in those with a positive cerebral angiogram (34% versus 13%, p = 0.0012), and a summary of the univariate analysis comparing those with positive and negative DSA results is shown in Table 3. Of note, patients presenting with HTN and pontine, cerebellar, and/or basal ganglia ICH (n = 111) had lower DSA positivity rates (9% versus 27%, p = 0.0002) compared to the rest of the patient sample.
Table 3.
Baseline patient characteristics across those with positive and negative DSA results.
| Variable a | DSA ( + ) n = 50 | DSA (−) n = 210 | p-value |
|---|---|---|---|
| Age, years | 44.06 (34–54) | 46.31 (39–56) | 0.36 |
| Male | 25 (50%) | 118 (56%) | 0.30 |
| Race |
0.42 |
||
| Caucasian | 30 (60%) | 119 (57%) | |
| African American | 15 (30%) | 76 (36%) | |
| Asian/other | 5 (10%) | 15 (7%) | |
| Location of ICH |
0.0081 |
||
| Cerebellum | 6 (12%) | 19 (9%) | |
| Pontine | 0 (0%) | 7 (3%) | |
| Basal Ganglia | 9 (18%) | 85 (40%) | |
| Cortical | 34 (68%) | 96 (46%) | |
| Other | 1 (2%) | 3 (1%) | |
| Smoking | 0.44 | ||
| Current or Former | 29 (48%) | 109 (52%) | |
| Never | 21 (42%) | 101 (48%) | |
| CTA obtained prior to DSA | 40 (80%) | 151 (72%) | 0.21 |
| Hypertensive b | 28 (56%) | 166 (79%) | 0.0013 |
| SAH | 17 (34%) | 27 (13%) | 0.0012 |
DSA: digital subtraction angiography; CTA: computed tomography angiography; ICH: intracranial hemorrhage; SAH: subarachnoid hemorrhage. The bold typeface indicates statistical significance.
Reported as mean (interquartile range) for continuous variables.
Hypertension was defined as having a systolic blood pressure >160 mmHg on presentation with ICH or a documented past medical history of hypertension as per the electronic medical record.
Of the 260 patients included in our study, 191 (73%) had a CTA study performed prior to diagnostic cerebral angiography. Of these 191 patients, 40 (21%) had a positive diagnostic cerebral angiography representing an underlying vascular lesion as the cause of the acute ICH. Therefore, we assessed the diagnostic accuracy of detecting an underlying vascular lesion between the two methodologies (Table 4). Overall, there was strong association between CTA and diagnostic cerebral angiogram results. CTA predicted the presence of an underlying vascular lesion with a robust specificity (94%) and sensitivity (97%). CTA failed to diagnose only one vascular lesion, which was a case of RCVS.
Table 4.
Comparison of CTA findings across patients with positive and negative DSA results.
| Variable | DSA ( + ) n = 40 | DSA (−) n = 151 | p-value |
|---|---|---|---|
| CTA findings | |||
| Positive | 27 (67.5%) | 5 (3%) | |
| Negative | 1 (2.5%) | 142 (94%) | < 0.0001 |
| Suspicious | 12 (30%) | 4 (3%) | |
| PPV: 81% NPV: 99% AUC: 0.96 Sensitivity: 97% Specificity: 94% | |||
DSA: digital subtraction angiography; CTA: computed tomography angiography. The bold typeface indicates statistical significance.
We developed a scoring system (ICH-Angio score) based on results of CTA (negative-0 points, suspicious-1 point, or positive-2 points), location of the ICH (basal ganglia/pons/cerebellum-0 points or cortical-1 point), history of HTN (yes-0 points or no-1 point), and presence of SAH (no-0 points or yes-1 point) to stratify patients into distinct risk groups based on the probability of having an underlying vascular lesion detected on diagnostic cerebral angiography. The ICH-Angio score criteria are presented in Table 5. We subsequently assigned the 191 patients with both CTA and cerebral angiography studies an ICH-Angio score based on the aforementioned criteria and assessed the diagnostic cerebral angiography positivity rates across the various risk groups. The DSA positivity rate for those with a score of 0 (0/62; 0%) was lower compared to those with a score of 1 (5/52; 10%), 2 (17/48; 35%), 3 (10/20; 50%), 4 (5/6; 83%), and 5 (3/3; 100%) which is illustrated in Figure 1. Approximately one-third (62/191, 32%) of patients were identified as ICH-Angio score of 0, suggesting a minimal risk for harboring an underlying vascular lesion. A linear-by-linear association test revealed a strong correlation between risk score and probability of having an underlying vascular lesion on DSA (p < 0.0001).
Table 5.
Proposed criteria for the ICH-Angio score.
| Variable | Score |
|---|---|
| CTA report | |
| Negative | 0 |
| Suspicious | 1 |
| Positive | 2 |
| Location | |
| Basal ganglia/pons/cerebellum | 0 |
| Cortical | 1 |
| Hypertension | |
| Yes (SBP > 160 mmHg or documented PMHx of HTN) | 0 |
| No | 1 |
| SAH | |
| Absent | 0 |
| Present | 1 |
CTA: computed tomography angiography; SBP: systolic blood pressure; PMHx: past medical history; HTN: hypertension; SAH: subarachnoid hemorrhage.
Figure 1.
Probability of a positive DSA exam stratified by ICH-Angio score.
The single case where CTA failed to diagnose the underlying vascular lesion of RCVS is presented in Figure 2. The patient was a 43-year old female with a history significant for selective serotonin reuptake inhibitor (SSRI) use who presented with HTN and multifocal hemorrhage with associated SAH detected on computed tomography of the head (CTH) (Figure 2(A) and (B)). CTA was negative for an underlying vascular lesion and the patient's risk score for positive DSA was 1 based on our aforementioned criteria. DSA demonstrated multifocal areas of stenosis in the internal, middle, and anterior cerebral artery distributions, consistent with a diagnosis of RCVS (Figure 2(C)).
Figure 2.
(A, B) A 43-year-old female presented headache, aphasia, and left-sided facial droop found to have bifrontal (right > left) ICHs with associated subarachnoid hemorrhage. CTA was negative for underlying lesion although (C) DSA demonstrated multifocal intracranial vessel narrowing in the anterior and middle cerebral artery distributions consistent with reversible cerebral vasoconstriction syndrome.
Discussion
Key results
Our study is the largest to assess the diagnostic yield of cerebral angiography in a population of young patients (≤ 60 years) presenting with acute ICH. Overall, our data suggests that age and HTN are negatively correlated with angiographic yield, and that non-invasive methods such as CTA can detect underlying vascular etiology of ICH with a robust specificity and sensitivity. Furthermore, we developed a novel scoring system based on CTA findings, location of ICH, history of HTN, and presence/absence of SAH to stratify patients based on risk for having an underlying vascular lesion, representing an effective way for the neurointerventionalist to quickly classify patients into risk groups for a vascular etiology of acute ICH. Using this system, patients without SAH, HTN, negative CTA, and non-lobar ICH were at a very low risk of harboring an underlying vascular lesion (0%).
Interpretation
A systematic review by Cordonnier et al. investigated the diagnostic utility of radiological techniques such as DSA to identify vascular etiologies underlying ICH. 7 Based on pooled results from 726 patients, DSA yield was lowest for those who presented with prestroke HTN and an ICH with a deep location. 7 However, a key limitation of that review was that it only considered AVMs and aneurysms as potential vascular etiologies, which may contribute towards an underestimate of the true diagnostic yield of DSA. In comparison, a particular strength of our study was the inclusion of several forms of vascular etiologies such as AVMs, aneurysms, Moyamoya disease, RCVS, and fistulas which may result in a more robust ability to capture the true diagnostic yield of DSA, as these pathologies are all potentially treatable. Furthermore, none of the studies included in that systematic review and meta-analysis focused exclusively on the younger patient population, which based on the trinational survey administered to 692 physicians in that same study, was the single most important factor influencing the decision to further investigate ICH. 7
Our results are in line with other reports such as that by Zhu et al. which showed that out of 148 normotensive patients, 66 (45%) had cerebral angiography findings positive for an underlying vascular etiology compared to only 9% (5/58) of those who were hypertensive. 5 However, there are noteworthy differences between that study and ours. That study was performed in a predominantly Asian population, and thus may not be generalizable to a more racially heterogeneous group of patients with unique predispositions to hypertensive and angiographic complications. The majority of patients in our study (194/260; 75%) were hypertensive whereas only 28% of the cohort in Zhu et al. was hypertensive. 5 This may partly explain why the diagnostic yield of cerebral angiography in patients with HTN was higher in our cohort compared to theirs (14% versus 9%). Furthermore, our data suggests that the diagnostic yield of cerebral angiography is higher for those with lobar ICHs, which is consistent with several other reports. For example, a retrospective study by Toffol et al. on 102 patients with nontraumatic ICH revealed a 47% diagnostic yield for lobar hemorrhages compared to 37.5% for non-lobar hemorrhages. 8 Furthermore, the diagnostic yield of angiography for non-lobar ICHs in hypertensive patients within that cohort was 0%. 8
Although our finding that vascular lesions are less likely to be detected in hypertensive patients with non-lobar ICH is consistent with previous studies, the majority of those reports cite a diagnostic yield of DSA close to 0%. In our cohort, the diagnostic yield of DSA for hypertensive patients with non-lobar ICH was slightly higher at 9%. This suggests that the rate of DSA positivity may be higher in an exclusively younger patient cohort. These patients may be effectively identified as harboring an underlying vascular etiology based on radiographic considerations such as associated SAH and CTA findings. For example, we presented a case example of a patient who presented with HTN, a multifocal ICH in non-lobar locations, and a negative CTA, but non-contrast CTH demonstrated a suspicious ICH pattern with associated SAH. DSA demonstrated findings consistent with a diagnosis of RCVS.
We assessed the diagnostic accuracy of CTA compared to cerebral angiography and found that CTA has excellent specificity (94%) and sensitivity (97%), suggesting that this non-invasive modality can be ubiquitously used as a screening tool to help rule in an underlying vascular etiology for acute ICH. This has been corroborated by other reports as well, suggesting that CTA may be nearly as effective as DSA for detecting underlying vascular lesions and revealing the etiology of acute ICHs. A study by Yeung et al. showed that CTA has excellent sensitivity (89%) and specificity (close to 92%) in identifying vascular malformations and etiologies of acute ICH. 9 Although the vast majority of lesions included in that study consisted of AVMs and aneurysms, more subtle disease processes such as Moyamoya disease and RCVS were assessed as well. 9 Furthermore, in our study, CTA detected all vascular etiologies of ICH, with the exception of one case of RCVS. It is known RCVS is a subtle disease process that is notoriously difficult to diagnose using non-invasive imaging such as CTA or MRA. 10 Although CTA may not be the most sensitive modality to detect subtle disease processes, other clinical and radiographic features may help direct clinicians toward obtaining a diagnostic cerebral angiogram on a case-by-case basis. For instance, in the illustrated RCVS case, the patient's clinical history of SSRI use and presence of SAH on CTH could guide clinicians to order further invasive imaging to establish the diagnosis.
We developed a novel scoring system (ICH-Angio score) based on CTA findings, location of ICH, medical history of HTN, and presence/absence of SAH to allow for rapid evaluation of a patient's risk for having a positive angiogram exam. Notably, we included the presence/absence of SAH as a criteria in the scoring system as it has been shown to have a strong association with vascular etiologies such as RCVS and ruptured aneurysms.11,12 We validated the ICH-Angio score within our cohort of young patients presenting with acute ICH, and show that it may have utility within the clinical setting by allowing for the rapid stratification of patients into risk groups. Notably, in our study, a young hypertensive patient with a negative CTA, absence of SAH, and a non-lobar ICH would receive a score of 0, which was associated with a 0% DSA-positivity rate. This is a clinically significant finding, as approximately one-third of patients were identified as having an ICH-Angio score 0, representing a significant group of patients who may be able to be evaluated using non-invasive imaging alone. In addition to guiding clinicians in ordering further invasive imaging, this framework may also be able to guide blood pressure management after spontaneous ICH, as lesions such as aneurysms and AVMs warrant aggressive blood pressure regulation. A report by Almandoz et al. similarly developed a scoring system to identify ICH patients with the highest risk of having an underlying vascular etiology. 13 Their proposed scoring system contained four criteria—age group, HTN or impaired coagulation history, and non-contrast CT categorization. 13 However, that study used Multidetector Computed Tomographic Angiography to make diagnoses which is a non-gold standard non-invasive imaging modality. 13 Furthermore, their cohort was substantially older than ours (mean age: 67 years, range: 18–94 years), and thus its utility in younger patients, for which determining whether cerebral angiography is warranted is more important, remains unclear. 13
Significance and generalizability
Knowledge of which ICH patients can be non-invasively identified as least likely to harbor an underlying vascular lesion can help reduce their exposure to the periprocedural risks associated with invasive neuroimaging techniques such as angiography. The ICH-Angio scoring system, which was validated on 191 patients, was able to non-invasively rule out an underlying vascular lesion in approximately one-third of patients. Our study, which is the first to exclusively focus on young ICH patients, revealed that the diagnostic yield of angiography in hypertensive patients with non-lobar ICH is substantially higher (10%) than reported in other studies. This finding underscores the importance of being vigilant towards underlying vascular etiologies of ICH, even when the pattern of ICH is consistent with a hypertensive etiology. Our ICH-Angio score considers CTA findings and the presence of SAH, which can assist neurointerventionalists in identifying this unique patient population for more definitive diagnostic imaging such as angiography. Furthermore, the results from our study are generalizable to an ethnically heterogeneous group of young ICH patients with increased predisposition for hypertensive and angiographic complications.
Limitations
We acknowledge that there are some limitations in our study. First, this is a single-center retrospective analysis, which limits the generalizability of our findings to other hospitals. Furthermore, we were unable to validate our scoring system on all 260 patients included in the study due to lack of both CTA and diagnostic cerebral angiography. Nonetheless, the majority (191/260; 73%) of our cohort underwent imaging by both modalities, reducing the risk that our findings regarding the scoring system would substantially change. Furthermore, there is likely some degree of inter-rater variability in classifying CTA results as either ‘suspicious’ versus ‘positive’. Large prospective multi-center trials are warranted to further elucidate the diagnostic utility of cerebral angiography for the subset of young patients presenting with acute ICH.
Conclusion
We validated a novel scoring system, the ICH-Angio score, to risk stratify spontaneous ICH patients for having an underlying vascular etiology. Within our cohort, approximately one-third of patients had a HTN history, negative CTA, absence of SAH, and a non-lobar ICH, which corresponded to a 0% risk of harboring an underlying vascular lesion. Non-invasive imaging modalities may be able to effectively screen patients for harboring underlying vascular causes for ICH that warrant further invasive angiographic evaluation.
Footnotes
Author contributions: AK and MDB were involved in the design and conception of this manuscript. MEE performed the literature search. MEE and AK compiled the primary manuscript. MEE and AK compiled the figures. RA, TP, NZM, and MDB critically revised the manuscript. All authors have approved the manuscript as it is written.
Data sharing: All data pertaining to this research article are included within the manuscript as written.
Mark D Bain is a consultant for CERENOVUS and Integra and a member of the advisory board for Stryker. All other authors have no personal or institutional interest with regards to the authorship and/or publication on this manuscript.
Funding: The author(s) received no financial support for the research, authorship, and/or publication of this article.
IRB approval: This study was approved by the Cleveland Clinic Institutional Review Board (Approval Number: 22-384).
ORCID iD: Mohamed E El-Abtah https://orcid.org/0000-0001-8451-7012
References
- 1.Caceres JA, Goldstein JN. Intracranial hemorrhage. Emerg Med Clin North Am 2012; 30: 771–794. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Fischbein NJ, Wijman CAC. Nontraumatic intracranial hemorrhage. Neuroimaging Clin N Am 2010; 20: 469–492. [DOI] [PubMed] [Google Scholar]
- 3.Hankey GJ, Warlow CP, Sellar RJ. Cerebral angiographic risk in mild cerebrovascular disease. Stroke 1990; 21: 209–222. [DOI] [PubMed] [Google Scholar]
- 4.Grzyska U, Freitag J, Zeumer H. Selective cerebral intraarterial DSA. Complication rate and control of risk factors. Neuroradiology 1990; 32: 296–299. [DOI] [PubMed] [Google Scholar]
- 5.Zhu XL, Chan MS, Poon WS. Spontaneous intracranial hemorrhage: which patients need diagnostic cerebral angiography? A prospective study of 206 cases and review of the literature. Stroke 1997; 28: 1406–1409. [DOI] [PubMed] [Google Scholar]
- 6.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. J Clin Epidemiol 2008; 61: 344–349. [DOI] [PubMed] [Google Scholar]
- 7.Cordonnier C, Klijn CJM, van Beijnum J, et al. Radiological investigation of spontaneous intracerebral hemorrhage: systematic review and trinational survey. Stroke 2010; 41: 685–690. [DOI] [PubMed] [Google Scholar]
- 8.Toffol GJ, Biller J, Adams HP, et al. The predicted value of arteriography in nontraumatic intracerebral hemorrhage. Stroke 1986; 17: 881–883. [DOI] [PubMed] [Google Scholar]
- 9.Yeung R, Ahmad T, Aviv RI, et al. Comparison of CTA to DSA in determining the etiology of spontaneous ICH. Can J Neurol Sci 2009; 36: 176–180. [DOI] [PubMed] [Google Scholar]
- 10.Grundt B, Bolling T, Ritch ML. Reversible cerebral vasoconstriction syndrome: a common occurrence but rare diagnosis. Cureus 2020; 12: e8546. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Petridis AK, Kamp MA, Cornelius JF, et al. Aneurysmal subarachnoid hemorrhage. Dtsch Arztebl Int 2017; 114: 226–236. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Muehlschlegel S, Kursun O, Topcuoglu MA, et al. Differentiating reversible cerebral vasoconstriction syndrome with subarachnoid hemorrhage from other causes of subarachnoid hemorrhage. JAMA Neurol 2013; 70: 1254–1260. [DOI] [PubMed] [Google Scholar]
- 13.Delgado Almandoz JE, Schaefer PW, Goldstein JN, et al. Practical scoring system for the identification of patients with intracerebral hemorrhage at highest risk of harboring an underlying vascular etiology: the secondary intracerebral hemorrhage score. AJNR Am J Neuroradiol 2010; 31: 1653–1660. [DOI] [PMC free article] [PubMed] [Google Scholar]