J NeuroIntervent Surg: first published as 10.1136/neurintsurg-2019-015375 on 4 November 2019. Downloaded from http://jnis.bmj.com/ on November 5, 2019 at Western Sydney University.
Neuroimaging
Department of Neurosurgery,
Barrow Neurological Institute,
Phoenix, Arizona, USA
Correspondence to
Dr Felipe C Albuquerque,
Department of Neurosurgery,
Barrow Neurological
Institute, Phoenix, AZ 85013,
USA; f​ elipe.​albuquerque@​
barrowbrainandspine.​com
Received 14 August 2019
Revised 4 October 2019
Accepted 18 October 2019
© Author(s) (or their
employer(s)) 2019. No
commercial re-­use. See rights
and permissions. Published
by BMJ.
To cite: Catapano JS,
Lang MJ, Koester SW, et al.
J NeuroIntervent Surg Epub
ahead of print: [please
include Day Month Year].
doi:10.1136/
neurintsurg-2019-015375
Original research
Digital subtraction cerebral angiography after
negative computed tomography angiography findings
in non-­traumatic subarachnoid hemorrhage
Joshua S Catapano, Michael J Lang, Stefan W Koester, Derrick J Wang,
Joseph D DiDomenico, Vance L Fredrickson, Tyler S Cole, Jonathan Lee,
Michael T Lawton, Andrew F Ducruet, Felipe C Albuquerque
Abstract
Background  CT angiography (CTA) is widely used for
the detection of vascular lesions in patients with non-­
traumatic subarachnoid hemorrhage (ntSAH); however,
digital subtraction angiography (DSA) remains the gold
standard for diagnosis. Our aim was to analyze the
diagnostic yield of DSA after negative high-­resolution
CTA findings.
Methods  Records of patients with a CTA-­negative
ntSAH at a single institution from 2014 to 2018
were retrospectively analyzed. ntSAH patterns were
categorized as cortical, perimesencephalic, or diffuse.
Subsequent DSA findings were compared across the
three cohorts.
Results  A total of 186 patients had CTA-­negative
ntSAH. The ntSAH pattern was identified as
cortical (n=77, 41.4%), diffuse (n=60, 32.3%), or
perimesencephalic (n=49, 26.3%). In eight patients
(4%), DSA revealed a vascular lesion (one cervical
arteriovenous fistula and seven atypical aneurysms) after
negative CTA findings. All eight patients with positive
DSA findings had diffuse SAH (13% of patients with
a diffuse pattern). The seven aneurysms included four
blister or dissecting (two basilar artery, one superior
cerebellar artery, and one dorsal wall internal carotid
artery), two fusiform (one posterior communicating
artery and one anterior spinal artery), and one saccular
aneurysm (middle cerebral artery).
Conclusion  DSA identified a causative lesion in
4% of patients with CTA-­negative ntSAH, but only in
patients with diffuse ntSAH. Most of the lesions detected
were atypical aneurysms and were found on delayed
angiograms. These results suggest that DSA can help
to diagnose CTA-­negative ntSAH caused by unusual
aneurysms, and repeat DSA may be needed only for
patients with diffuse ntSAH.
Introduction
Non-­traumatic subarachnoid hemorrhage (ntSAH)
occurs in approximately 30 000 patients per year,
or 5% of strokes in the USA.1 Although most
SAHs are caused by rupture of a brain aneurysm,
a causative lesion is not found in 15–20% of
patients.2–4 Patients with aneurysmal SAH experi-
ence high rates of morbidity and mortality; there-
fore, prompt diagnosis of a ruptured aneurysm is
essential.
Catapano JS, et al. J NeuroIntervent Surg 2019;0:1–5. doi:10.1136/neurintsurg-2019-015375
Historically, digital subtraction angiography
(DSA) has served as the gold standard for diagnosis,
with a reported sensitivity of 99%.5 6 However, with
the advent of higher-­resolution CT technologies,
CT angiography (CTA) is increasingly used, often
displacing DSA as the first measure.7 Although CTA
is notably less expensive, less invasive, and more
widely available,8 it also exhibits a lower sensitivity
for the detection of intracranial aneurysms,9–12
failing to identify a causative lesion in 5–30% of
cases.13
In cases with negative CTA findings, DSA can
identify a causative lesion in 4–14% of patients
Protected by copyright.
with SAH, and the diagnostic yield of repeat DSA
ranges from 4% to 16%.13 Review of the literature,
however, primarily reports findings obtained using
outdated CTA technology. Our aim was to analyze
the diagnostic yield of DSA after high-­resolution
modern CTA with negative findings during a 5-­year
period at our high-­volume neurovascular center.
Methods
At a single tertiary care center, medical records
from a prospectively maintained neuroendovascular
database were reviewed for patients with an initial
presentation of ntSAH and negative CTA findings
who underwent subsequent DSA between January
1, 2014 and December 31, 2018. The study was
approved by the institutional review board of our
hospital and complied with the Health Insurance
Portability and Accountability Act (HIPAA). Images
were obtained with use of a GE LightSpeed VCT
CT scanner using a 64-­slice CTA protocol for high
resolution (GE Healthcare, Chicago, Illinois, USA).
Hospital notes, age, sex, and imaging studies (CT,
CTA, and DSA) were analyzed. The pattern of ntSAH
was categorized as diffuse (figure 1A), perimesen-
cephalic (figure 1B), or cortical (figure 1C) on the
basis of a review of the report by a cerebrovascular
neurosurgery attending physician. Cortical ntSAH
was defined as that occurring near the convexity,
without the involvement of the basilar cisterns.
Perimesencephalic ntSAH was defined using previ-
ously described features.14 In general, this pattern is
localized to the cisterns at the level of the midbrain,
often centered around the interpeduncular cistern.
Additional extension, particularly superior to this
location and with blood layering in multiple cisterns
and subarachnoid spaces, was defined as diffuse
   1
 J NeuroIntervent Surg: first published as 10.1136/neurintsurg-2019-015375 on 4 November 2019. Downloaded from http://jnis.bmj.com/ on November 5, 2019 at Western Sydney University.
Neuroimaging

Figure 1  CT of the head without contrast. (A) Diffuse non-­traumatic subarachnoid hemorrhage (ntSAH) is diffusely layered in multiple cisterns and
subarachnoid spaces. (B) Perimesencephalic ntSAH is localized to the cisterns at the level of the midbrain, centered often around the interpeduncular
cistern. (C) Cortical ntSAH occurs near the convexity, without involvement of the basilar cisterns. Used with permission from Barrow Neurological
Institute, Phoenix, Arizona, USA.

ntSAH. Negative CTA findings were defined as no lesion being
identified or suspected on the final attending neuroradiologist
report (seven patients had suspected vasculitis as the cause for
ntSAH and were excluded).
If the first DSA after a CTA with negative findings was unre-
rate for perimesencephalic, cortical, and diffuse hemorrhage on
CTA was 0%, 0%, and 7.3%, respectively. Table 3 highlights
recent large studies on CTA-­negative ntSAH, with the majority
of lesions found among patients with diffuse ntSAH.7 13 15–17

vealing, a follow-­up DSA was performed at the discretion of the
neuroendovascular attending physician, generally 5–7 days after
the initial study. Statistical analysis was performed using IBM
SPSS Statistics for Windows Version 24.0 (IBM Corp, Armonk,
New York, USA) and Microsoft Excel (Microsoft, Redmond,
Washington, USA).
Protected by copyright.
Discussion
CTA has emerged as the first-­line diagnostic study for patients
presenting with SAH, as it is more sensitive and less time-­
consuming than magnetic resonance angiography and less inva-
sive than DSA.18 19 However, in 5–30% of patients with SAH,
CTA does not identify a causative vascular lesion.20 21 Because

Results
From January 1, 2014 to December 31, 2018, 186 patients Table 1  Characteristics of patients with CTA-­negative ntSAH
presented with ntSAH and had negative initial CTA findings. Patients
Of these patients, 99 were men (53%) and the mean (SD) age Characteristic (n=186)
was 56 (14) years (table 1). The SAH pattern was identified as Sex
cortical (n=77, 41.4%), diffuse (n=60, 32.3%), or perimesen-  Men 97 (52)
cephalic (n=49, 26.3%). Intraventricular or intraparenchymal
 Women 89 (48)
hemorrhage was present in 40 (21.5%) and 18 (9.7%) patients,
respectively. Age, mean (SD), years 56 (14)
Eight patients (4%) had a DSA positive for a vascular lesion ntSAH pattern
(one cervical arteriovenous fistula (AVF) and seven atypical  Diffuse 60 (32)
aneurysms) after a CTA with negative findings (table 2). All eight
 Perimesencephalic 49 (26)
patients with positive DSA findings had a diffuse pattern of SAH
(13% of patients with a diffuse pattern). Of the seven aneurysms,  Sulcal 77 (41)
four were blister or dissecting (two basilar artery, one superior  IVH 40 (21)
cerebellar artery, and one dorsal wall internal carotid artery),  IPH 18 (10)
two were fusiform (one posterior communicating artery, diam-
Repeat DSA
eter 2 mm, and one anterior spinal artery), and one was saccular
(middle cerebral artery, diameter 1 mm).  2 DSAs 109 (59)
All 186 patients underwent initial DSA at a mean of <1 day  3 DSAs 18 (10)
from admission (median 1 day; range 0–8 days), and 109 patients Vascular lesions found
underwent repeat DSA at a mean of 7.6 days from the first study
 Overall (n=186) 8 (4.3)
(median 6 days; range 2–9 days; one outlier at 148 days). A caus-
ative lesion was identified in two additional patients at the time  First DSA (n=186) 4 (2.2)
of repeat DSA (two aneurysms; one fusiform, one dissecting).  Second DSA (n=109) 2 (1.8)
Eighteen patients underwent a third DSA at a mean of 110 days  Third DSA (n=18) 2 (11.1)
from the second DSA (median 75 days; range 1–416 days). The Data are n (%) of patients unless otherwise indicated.
third study identified a causative lesion in two patients (two aneu- CTA, CT angiography; DSA, digital subtraction angiography; IPH, intraparenchymal;
rysms; one fusiform, one dissecting). The overall false-­negative IVH, intraventricular; ntSAH, non-­traumatic subarachnoid hemorrhage.

2 Catapano JS, et al. J NeuroIntervent Surg 2019;0:1–5. doi:10.1136/neurintsurg-2019-015375

 J NeuroIntervent Surg: first published as 10.1136/neurintsurg-2019-015375 on 4 November 2019. Downloaded from http://jnis.bmj.com/ on November 5, 2019 at Western Sydney University.
Neuroimaging

Table 2  Patients with CTA-­negative ntSAH and vascular lesion found on DSA
No of DSAs with
negative
Patient findings SAH type Lesion Type Location
1 0 Diffuse Aneurysm Blister Dorsal wall ICA (figure 2A)
2 0 Diffuse AVF N/A Cervical (figure 2H)
3 0 Diffuse Aneurysm Dissecting Basilar (figure 2C)
4 0 Diffuse Aneurysm Saccular MCA (diameter 1 mm) (figure 2D)
5 1 Diffuse Aneurysm Dissecting Basilar apex (figure 2B)
6 1 Diffuse Aneurysm Fusiform PCOM (diameter 2 mm) (figure 2F)
7 2 Diffuse Aneurysm Fusiform ASA (figure 2E)
8 2 Diffuse Aneurysm Dissecting SCA (figure 2G)
ASA, anterior spinal artery; AVF, arteriovenous fistula; CTA, CT angiography; DSA, digital subtraction angiography; ICA, internal carotid artery; MCA, middle cerebral artery; ntSAH,
non-­traumatic subarachnoid hemorrhage; PCOM, posterior communicating artery; SCA, superior cerebellar artery.

of the devastating consequences of a false-­negative finding in the
diagnosis of a ruptured vascular lesion, most of these patients are
subsequently evaluated by DSA.22
In our study, DSA revealed a cause of the hemorrhage in eight
(4%) patients with CTA-­negative ntSAH, all of whom presented
with a diffuse pattern of SAH. Seven vascular lesions detected by
DSA were atypical ruptured aneurysms and one was a cervical
AVF (figure 2A–H). These results are similar to previous retro-
spective analyses in which aneurysms were the primary causative
vascular lesion in patients with CTA-­ negative, DSA-­ positive
atypical and dangerous features. Similar diagnostic yields for
finding uncommon or small aneurysms have been reported in
the literature for repeat DSAs.24
In our study, DSA did not reveal a causative vascular lesion
for any of the cases with a perimesencephalic or cortical pattern
of SAH. Although, in more recent studies (table 3), the majority
of lesions in CTA-­negative ntSAH are found among patients
with diffuse SAH, some lesions are identified among patients
with other bleed patterns, most of which were caused by vascu-
litis.7 13 15–17 Historically, perimesencephalic SAH has been asso-

diffuse SAH.7 16 Less frequently, arteriovenous malformations
or intracranial dural AVFs were identified as the source of
bleeding.13 16 In a 2017 retrospective cohort study by Philipp
et al,23 the sensitivity of CTA was notably only 57.6% for aneu-
rysms smaller than 5 mm in diameter. As such, small aneurysms
may be missed in patients undergoing CTA alone.
For cases of diffuse SAH with both negative CTA and negative
DSA findings, a vascular lesion still may be the cause. Negative
DSA findings in patients with SAH caused by a vascular lesion
may be due to adjacent blood causing compression of the aneu-
rysm, aneurysmal thrombosis, vasospasm of the parent vessel,
or small aneurysm size.22 In our study, we report the results of
additional follow-­up DSAs in the event of negative CTA find-
ings and negative baseline DSA findings. With this protocol,
ruptured aneurysms were found in two of 109 patients (1.8%)
on a second DSA and in two of 18 patients (11.1%) on a third
DSA. Notably, all of these lesions were small and possessed
Protected by copyright.
ciated with negative DSA findings.4 7 16 25 Approximately 10% of
all spontaneous SAH cases exhibit this pattern of hemorrhage,
with venous bleeding as the suspected source. Nonetheless, addi-
tional studies have reported cases of false-­negative CTA findings
for perimesencephalic SAH and source identification in 3–4.5%
of cases after DSA.13 24 As a result, an initial DSA is supported
for perimesencephalic SAH, despite the expected low diagnostic
yield. However, in the event of an initial DSA negative for the
perimesencephalic pattern, our results are in agreement with a
report by Huttner et al26 in which additional DSAs did not yield
further evidence of a vascular lesion.
Timing of DSA may play a role in identifying vascular lesions.
The median time interval between the initial CTA and the first
DSA was 1 day in our series. The median time between the first
DSA and the second DSA was 6 days, followed by a median of
75 days to the third DSA. These time intervals between studies
are shorter than those in the existing literature. In particular,

Table 3  Major studies on patients with CTA-­negative ntSAH and DSA findings, 2006–2018
No of patients with CTA-­
negative non-­diffuse
No (%) of patients ntSAH with vascular
No of patients with with DSA-­positive, lesion on DSA (% of
Study, year Period of review Study type CT scanner(s) used CTA-­negative ntSAH CTA-­negative ntSAH lesions)
Bashir et al, 201815 2011–2017 Retrospective case series Unknown 74* 1 (1.4) 0
Heit et al, 201613 2002–2012 Retrospective case series 64 slice and 16 slice 230 29 (13) 14 (48)†
Delgado Almandoz et al, 2007–2011 Retrospective case series 64 slice and 16 slice 55‡ 6 (11) 1 (17)
201316
Agid et al, 20107 2005–2009 Retrospective case series 64 slice 193 12 (6) 7 (58)§
Jung et al, 200617 1986–2004 Retrospective case series Unknown 143 18(18) Unknown
*Study included combined CTA-­negative and DSA-­negative ntSAH.
†Twelve negative-­CTA ntSAH due to vasculitis.
‡Magnetic resonance angiography used for eight patients.
§All non-­diffuse ntSAH with a lesion found on DSA were due to vasculitis.
CTA, CT angiography; DSA, digital subtraction angiography; ntSAH, non-­traumatic subarachnoid hemorrhage.

Catapano JS, et al. J NeuroIntervent Surg 2019;0:1–5. doi:10.1136/neurintsurg-2019-015375 3

 J NeuroIntervent Surg: first published as 10.1136/neurintsurg-2019-015375 on 4 November 2019. Downloaded from http://jnis.bmj.com/ on November 5, 2019 at Western Sydney University.
Neuroimaging

Protected by copyright.
Figure 2  Vascular lesions identified on digital subtraction angiography among patients with CT angiography (CTA)-­negative non-­traumatic
subarachnoid hemorrhage. (A) Reconstructed three-­dimensional (3D) image of a dorsal wall right internal carotid artery blister aneurysm (arrow)
that was not seen on the initial CTA due to the small size and abnormal location of the lesion. (B) Reconstructed 3D image of a dissecting basilar
apex aneurysm (arrow) that was not visualized on the initial CTA due to the small size of the lesion. (C) Reconstructed 3D image of dissecting basilar
artery aneurysm (arrow) that was not identified on the initial CTA due to the small size of the lesion. (D) Reconstructed 3D image of a small 1 mm
right middle cerebral artery aneurysm (arrow) that was not seen on the initial CTA due to the small size of the lesion. (E) Right vertebral artery
injection with a lateral working view showing a small fusiform atrial septal aneurysm (arrow) that was not visible on the initial CTA because the
aneurysm was too caudal. (F) Lateral view of a fusiform left posterior communicating artery aneurysm (arrow) not seen on the initial CTA because a
large subarachnoid hemorrhage around the lesion made the aneurysm difficult to visualize. (G) Right vertebral artery injection posterioranterior view
showing a dissecting left superior cerebellar artery aneurysm (arrow) that was not visualized on the initial CTA due to the small size of the lesion
and a large subarachnoid hemorrhage around the lesion. (H) Lateral view of a left vertebral artery injection showing a cervical arteriovenous fistula
(arrow) that was not visible on the initial CTA because the fistula was too caudal. Used with permission from Barrow Neurological Institute, Phoenix,
Arizona, USA.

reports vary in the interval between CTA and DSA, ranging from
less than 1 day to 7 days.24 26 27 As the diagnostic yield from
DSA relative to the timing of the procedure has varied between
studies, no clear association can be made between timing and the
discovery of a causative vascular lesion.
Limitations of our study include those inherent to a retrospec-
tive analysis. Furthermore, the definition of SAH patterns may
differ among papers and institutions, leading to variable results.
In addition, repeat DSAs were ordered on the basis of the prefer-
ences of the attending neuroendovascular surgeon. In the future,
a prospective study design with standardized SAH diagnostic
criteria and DSA repeat time intervals could help the develop-
ment of concise guidelines.
Conclusion
DSA identified a causative vascular lesion in 4% of patients
with CTA-­negative ntSAH, even with the use of high-­resolution
modern-­era CTA. A vascular lesion was identified by DSA only
in patients with diffuse pattern ntSAH. Furthermore, most of the
vascular lesions detected were atypical aneurysms, and half were
found on the second or third DSA study. These results support
the importance of performing DSA for patients with CTA-­
negative ntSAH to diagnose atypical or small aneurysms, with
repeat DSAs for patients with diffuse pattern ntSAH.
Acknowledgements  The authors thank the Neuroscience Publications staff at
Barrow Neurological Institute for help with manuscript preparation.

4 Catapano JS, et al. J NeuroIntervent Surg 2019;0:1–5. doi:10.1136/neurintsurg-2019-015375


Contributors  JSC: study design, data collection, data analysis, and manuscript
editing and writing. ML: study design and data analysis. SK: data collection and
manuscript writing. DW: data collection. JDiD: manuscript editing and data analysis.
VLF: manuscript editing and data analysis. TSC: manuscript editing and data analysis.
JL: manuscript writing. MTL: study design and manuscript editing. AD: study design
and manuscript editing. FCA: study design, manuscript editing, and guarantor.
Funding  The authors have not declared a specific grant for this research from any
funding agency in the public, commercial, or not-­for-­profit sectors.
Competing interests  None declared.
Patient consent for publication  Not required.
Provenance and peer review  Not commissioned; externally peer reviewed.
Data availability statement  Data are available upon reasonable request.
References
1 Bederson JB, Connolly ES, Batjer HH, et al. Guidelines for the management of
aneurysmal subarachnoid hemorrhage: a statement for healthcare professionals from
a special writing group of the Stroke Council, American Heart Association. Stroke
2009;40:994–1025.
2 Duong H, Melançon D, Tampieri D, et al. The negative angiogram in subarachnoid
haemorrhage. Neuroradiology 1996;38:15–19.
3 Kaim A, Proske M, Kirsch E, et al. Value of repeat-­angiography in cases of unexplained
subarachnoid hemorrhage (SAH). Acta Neurol Scand 1996;93:366–73.
4 Rinkel GJ, Wijdicks EF, Hasan D, et al. Outcome in patients with subarachnoid
haemorrhage and negative angiography according to pattern of haemorrhage on
computed tomography. Lancet 1991;338:964–8.
5 Luo Z, Wang D, Sun X, et al. Comparison of the accuracy of subtraction CT
angiography performed on 320-­detector row volume CT with conventional CT
angiography for diagnosis of intracranial aneurysms. Eur J Radiol 2012;81:118–22.
6 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;36:176–80.
7 Agid R, Andersson T, Almqvist H, et al. Negative CT angiography findings in patients
with spontaneous subarachnoid hemorrhage: when is digital subtraction angiography
still needed? AJNR Am J Neuroradiol 2010;31:696–705.
8 Hoh BL, Cheung AC, Rabinov JD, et al. Results of a prospective protocol of computed
tomographic angiography in place of catheter angiography as the only diagnostic and
pretreatment planning study for cerebral aneurysms by a combined neurovascular
team. Neurosurgery 2004;54:1329–42.
9 Fox AJ, Symons SP, Aviv RI. CT angiography is state-­of-­the-­art first vascular imaging
for subarachnoid hemorrhage. AJNR Am J Neuroradiol 2008;29:e41–2.
10 Kelliny M, Maeder P, Binaghi S, et al. Cerebral aneurysm exclusion by CT angiography
based on subarachnoid hemorrhage pattern: a retrospective study. BMC Neurol
2011;11:8.
Catapano JS, et al. J NeuroIntervent Surg 2019;0:1–5. doi:10.1136/neurintsurg-2019-015375
J NeuroIntervent Surg: first published as 10.1136/neurintsurg-2019-015375 on 4 November 2019. Downloaded from http://jnis.bmj.com/ on November 5, 2019 at Western Sydney University.
Neuroimaging
11 Kershenovich A, Rappaport ZH, Maimon S. Brain computed tomography angiographic
scans as the sole diagnostic examination for excluding aneurysms in patients with
perimesencephalic subarachnoid hemorrhage. Neurosurgery 2006;59:798–802.
12 Prestigiacomo CJ, Sabit A, He W, et al. Three dimensional CT angiography versus
digital subtraction angiography in the detection of intracranial aneurysms in
subarachnoid hemorrhage. J Neurointerv Surg 2010;2:385–9.
13 Heit JJ, Pastena GT, Nogueira RG, et al. Cerebral angiography for evaluation of
patients with CT angiogram-­negative subarachnoid hemorrhage: an 11-­year
experience. AJNR Am J Neuroradiol 2016;37:297–304.
14 Rinkel GJ, Wijdicks EF, Vermeulen M, et al. Nonaneurysmal perimesencephalic
subarachnoid hemorrhage: CT and MR patterns that differ from aneurysmal rupture.
AJNR Am J Neuroradiol 1991;12:829–34.
15 Bashir A, Mikkelsen R, Sørensen L, et al. Non-­aneurysmal subarachnoid hemorrhage:
when is a second angiography indicated? Neuroradiol J 2018;31:244–52.
16 Delgado Almandoz JE, Crandall BM, Fease JL, et al. Diagnostic yield of catheter
angiography in patients with subarachnoid hemorrhage and negative initial
noninvasive neurovascular examinations. AJNR Am J Neuroradiol 2013;34:833–9.
17 Jung JY, Kim YB, Lee JW, et al. Spontaneous subarachnoid haemorrhage with negative
initial angiography: a review of 143 cases. J Clin Neurosci 2006;13:1011–7.
18 Anderson GB, Steinke DE, Petruk KC, et al. Computed tomographic angiography
versus digital subtraction angiography for the diagnosis and early treatment of
ruptured intracranial aneurysms. Neurosurgery 1999;45:1315–22.
19 Velthuis BK, Van Leeuwen MS, Witkamp TD, et al. Computerized tomography
angiography in patients with subarachnoid hemorrhage: from aneurysm detection to
treatment without conventional angiography. J Neurosurg 1999;91:761–7.
20 Bradac GB, Bergui M, Ferrio MF, et al. False-­negative angiograms in subarachnoid
haemorrhage due to intracranial aneurysms. Neuroradiology 1997;39:772–6.
21 Urbach H, Zentner J, Solymosi L. The need for repeat angiography in subarachnoid
haemorrhage. Neuroradiology 1998;40:6–10.
22 van Gijn J, Rinkel GJ. Subarachnoid haemorrhage: diagnosis, causes and management.
Brain 2001;124:249–78.
23 Philipp LR, McCracken DJ, McCracken CE, et al. Comparison between CTA and digital
subtraction angiography in the diagnosis of ruptured aneurysms. Neurosurgery
2017;80:769–77.
Protected by copyright.
24 Delgado Almandoz JE, Jagadeesan BD, Refai D, et al. Diagnostic yield of repeat
catheter angiography in patients with catheter and computed tomography
angiography negative subarachnoid hemorrhage. Neurosurgery 2012;70:1135–42.
25 van Gijn J, van Dongen KJ, Vermeulen M, et al. Perimesencephalic hemorrhage:
a nonaneurysmal and benign form of subarachnoid hemorrhage. Neurology
1985;35:493–7.
26 Huttner HB, Hartmann M, Köhrmann M, et al. Repeated digital substraction
angiography after perimesencephalic subarachnoid hemorrhage? J Neuroradiol
2006;33:87–9.
27 Hijdra A, Braakman R, van Gijn J, et al. Aneurysmal subarachnoid hemorrhage:
complications and outcome in a hospital population. Stroke 1987;18:1061–7.
5