Yield of angiographic examinations in isolated intraventricular hemorrhage: A case series and systematic review of the literature

Abstract

Background

It is unknown which patients with non-traumatic isolated intraventricular hemorrhage should undergo angiographic imaging to detect an underlying macrovascular cause and which modality has the highest yield. We studied yield of angiographic examinations in patients with isolated intraventricular hemorrhage.

Methods

We reviewed medical records of patients with intraventricular hemorrhage admitted to the University Medical Center Utrecht between 2002 and 2012. We searched PubMed and Embase for studies on angiographic examinations in intraventricular hemorrhage until January 2014. We calculated yield of angiographic imaging and investigated influence of age, hypertension and anticoagulant use with meta-regression analysis.

Results

We identified 39 patients of whom 30 underwent an angiographic study. CTA suggested a macrovascular abnormality in nine patients, which was confirmed by DSA in seven. In the literature, we found 16 studies describing 209 patients. Pooled analysis showed a yield of 58% for DSA (95% CI 48–68%; 147 patients). One small study described the yield of CTA or MRA (0%; 4 patients). Yield of angiographic imaging decreased with increasing age (−2.6%; −5.0 to −0.2 per year increase) but was not affected by history of hypertension (−8.3%; −80.8 to 64.2) or anticoagulant use (−47.1%; −110.3 to 16.1).

Conclusion

The reported yield of DSA in isolated intraventricular hemorrhage is around 50% but varies considerably, probably due to differences in clinical judgment on the need for angiography performance. The yield is higher in younger patients but based on the available data, it is not possible to set age or other criteria for patients in whom DSA can be safely omitted.

Keywords

Intraventricular hemorrhage digital subtraction angiography computed tomography angiography magnetic resonance angiography arteriovenous malformation

Introduction

Isolated intraventricular hemorrhage (IVH) accounts for around 3% of all non-traumatic intracranial hemorrhages in adults and is therefore a rare condition.^1,2 Isolated IVH can originate from rupture of an underlying macrovascular abnormality, e.g. an aneurysm, arteriovenous malformation (AVM) or dural arteriovenous fistula (dAVF). Identification of such abnormalities has important therapeutic implications. Digital subtraction angiography (DSA) is the gold standard for detection of underlying macrovascular abnormalities, but less invasive computed tomography angiography (CTA) and magnetic resonance angiography (MRA) are increasingly applied. It is unclear in which patients with IVH angiographic studies should be performed and what the diagnostic accuracy is of non-invasive techniques. Also, yield of angiographic imaging might be influenced by patient characteristics, including age, pre-existing hypertension and use of oral anticoagulants.^3,4 The purpose of this study was to investigate the yield of angiographic modalities in patients with isolated IVH and to study the influence of patient characteristics on this yield.

Methods

Case series

We reviewed the stroke database of the University Medical Center Utrecht for consecutive patients who had been admitted with isolated IVH between January 2002 and May 2012. We reviewed patient records for information on age, sex, diagnosis of hypertension and use of anticoagulants. Patients were considered hypertensive when they had a medical history of hypertension, used antihypertensive agents on admission or were discharged on antihypertensive medication. We reviewed non-contrast CT scans for signs of small vessel disease (SVD), which was defined as the presence of white matter lesions, or an infarct in basal ganglia, thalamus or posterior fossa. We extracted the number, type and yield of angiographic examinations.

Systematic literature search

We systematically searched PubMed and Embase up to January 2014 using synonyms of the following search terms ‘intraventricular hemorrhage’, ‘digital subtraction angiography’, ‘magnetic resonance angiography’ and ‘computed tomography angiography’ (NH, Supplementary Table 1). No language restrictions were added. Title and abstract were screened for relevant articles based on predefined inclusion criteria. Inclusion criteria were (1) IVH defined as non-traumatic hemorrhage confined to the ventricles, (2) inclusion of at least three consecutive patients, (3) reporting results of CTA, MRA or DSA. References and related citations of relevant articles were screened until no additional studies were found. The quality of the selected studies was assessed using the Quality Assessment of Diagnostic Accuracy Studies (QUADAS),⁵ and the Standards for the Reporting of Diagnostic accuracy studies (STARD).⁶

Two authors (NH, CvA) independently retrieved information on case-finding methods, the number of included patients, age, sex, history of hypertension, use of anticoagulants, the applied angiographic modalities and their diagnostic yield. Diagnostic yield was defined as the proportion of angiographic examinations that identified a vascular abnormality as the cause of the IVH.

Statistical analysis

For each study, we calculated mean age, proportion of men, proportion of patients with hypertension, proportion of patients on anticoagulants and the yield of angiographic examinations with 95% CI. In the case series, we calculated sensitivity and specificity of CTA compared with DSA, with 95% CI. Inconclusive CTAs were considered as positive outcomes. We performed logistic regression to study the association between age, signs of SVD and presence of vascular malformations. We pooled the yield of angiographic examinations obtained from the case series with the estimates from the literature review using a random effects model. To investigate the association between age, hypertension, use of anticoagulants, and the yield, we performed meta-regression analysis weighted by the inverse variance. All statistical analyses were performed with SPSS version 22 or R version 3.0.3 for Windows (http://cran.r-project.org/).

Results

Case series

In our case series, we included 39 patients with isolated IVH (Supplementary Table 2). In 30 patients, one or more angiographic examination had been performed: CTA in 28 patients, MRA in five patients and DSA in nine. In seven of 30 patients (23%), a macrovascular cause was found, in five patients an AVM, and in two patients a dAVF. CTA had suggested the underlying macrovascular abnormality in nine patients, and it was confirmed by DSA in seven patients. Sensitivity of CTA was 100% (95% CI 56–100) and specificity 50% (95% CI 3–97%). No DSA was performed in any of the 18 patients with a negative CTA. In the five patients who underwent MRA, we found no underlying macrovascular lesions; in three of these patients, prior CTA had been negative as well. During a median follow-up of three months (range 0–132 months), we found no macrovascular abnormalities nor recurrent hemorrhage among 24 patients with non-fatal IVH, who did not undergo DSA. The odds of having a vascular malformation decreased with each year increase in age (OR 0.91, 95% CI 0.85–0.97). There was no significant association between signs of SVD and presence of a macrovascular abnormality (OR 0.45, 95% CI 0.08–2.69).

Systematic literature search

The literature search yielded 16 studies reporting on 209 patients with isolated IVH (Supplementary Figure 1).^1,2,4,7–19 Thirteen studies described the yield of a single investigation, in all cases DSA. One study described DSA and performed MRA in one patient,¹¹ one study compared CTA with DSA,¹⁹ and a single study reported on the diagnostic yield of CTA in addition to gadolinium-enhanced MRI/MRA.⁷ Six studies had a prospective design.^{1,7,8,10,16,17} All studies except two based the diagnosis isolated IVH on initial CT imaging. Two studies also performed MRI, either in a subset of patients,¹¹ or in all patients.⁷ Four studies specifically state that CT scans were reviewed by at least two radiologists.^7,8,10,11 None of the included studies fulfilled all criteria for good quality of diagnostic studies according to the QUADAS-tool or STARD-checklist.^5,6 Apart from two, none of the studies assessed the yield of one modality in comparison with another that was used as a reference standard,^7,19 and most studies did not specify their selection criteria for angiography (Supplementary Figure 2).

Baseline characteristics of studies are summarized and combined with the results of our case series in Table 1. Mean age was reported in eight studies and ranged from 46 to 61 years.^{1,4,7,9,12,13,17} Median proportion of men was 58% (range 38–80%, 10 studies).^{1,4,7,9,11–13,17,19} Median proportion of patients with hypertension was 50% (range 0–80%, 10 studies).^1,4,9–15 Yield of DSA in 16 studies ranged from 29 to 100%; a pooled analysis showed a yield of 58% (95% CI 48 to 68%) for DSA (Figure 1; I² 39%). Yield of CTA compared with DSA, as assessed in one study, was not reported separately for patients with isolated IVH and authors were unable to provide additional information after request.¹⁹ Yield of MRA and CTA, investigated in a study with four consecutive patients, was 0%.⁷

Table 1.

Characteristics of 17 cohorts reporting on the yield of angiographic examinations in isolated intraventricular hemorrhage, including our own series.

Study (first author, year of publication)	Study design	No. of patients (n)	Mean age	Male (%)	Hypertension (%)	Use of OAC (%)	Angiographic modality	No. of patients undergoing angiographic studies	Macrovascular cause found (%, 95% CI)
Hilkens et al., 2016	R	39	58	59	46	36	CTA-MRA-DSA	28–5–9	23 (10–42)
Snider et al., 2010⁷	P	4	61	50	–	–	CTA-MRA	4–4	0 (0–60)
Giray et al., 2009⁴	R	24	61	58	50	4	DSA	17	29 (10–56)
Yeung et al., 2009¹⁹	R	5	median 49 (m) median 50 (f)	58	–	–	CTA-DSA	5–5	60 (15–95)
Flint et al., 2008⁹	R	15	60	47	80	7	DSA	9	44 (14–79)
Abu Bakar et al., 2005⁸	P	3	–	–	–	–	DSA	3	67 (9–99)
Hameed et al., 2005¹¹	R	15	median 56	60	80	13	DSA-MRA	3–1	50 (7–93)
Passero et al., 2002¹	P	26	58	50	38	4	DSA	18	44 (22–69)
Marti-Fabregas et al., 1999¹³	R	13	60	38	54	0	DSA	5	40 (5–85)
Lin and Howng, 1997¹⁸	R	8	15–40	–	–	–	DSA	8	75 (35–97)
Zhu et al., 1997¹⁶	P	17	–	–	–	–	DSA	17	65 (38–86)
Roos et al., 1995²	R	24	–	–	–	–	DSA	14	57 (29–82)
Halpin et al., 1994¹⁰	P	3	under 40	–	0	–	DSA	3	67 (9–99)
Weisberg et al., 1991¹⁵	R	26	–	–	50	–	DSA	26	42 (23–63)
Jayakumar et al., 1989¹²	R	15	45	80	33	–	DSA	12	42 (15–72)
Darby et al., 1988¹⁷	P	5	46	60	–	–	DSA	3	100 (29–100)
Toffol et al., 1986¹⁴	R^a	6	–	–	50	–	DSA	6	83 (36–100)

Patients were selected by reviewing angiographic records. P = prospective; R = retrospective; m = male; f = female; OAC = oral anticoagulants.

Figure 1.

Pooled analysis of 16 studies reporting on yield of digital subtraction angiography.

Overall, 84 macrovascular abnormalities were identified: in 50 patients (60%) the cause of IVH was an AVM, in 28 (35%) an aneurysm, in five (6%) a dural AVF and in one patient moyamoya disease (1%). With each year increase in mean age per study, the yield of angiographic imaging decreased by −2.6% (95% CI −5.0 to −0.2). We found no effect of the proportion of patients with hypertension (−8.3%; 95% CI −80.8 to 64.2) or of patients on anticoagulants (−47.1%; 95% CI −110.3 to 16.1).

Discussion

In around half of the patients with IVH undergoing DSA, a macrovascular lesion is found, but the yield varies considerably between studies, which is probably explained by varying selection criteria. The yield of angiographic imaging increases in younger patients, whereas we found no effect of the presence of hypertension or the use of oral anticoagulants.

Accurate detection of underlying malformations is of great importance, since timely treatment may prevent recurrent hemorrhage. Although the yield was higher in younger patients, we could not set an age criterion for patients in whom DSA can be omitted based on our data; it is likely that also here bias by indication played a role in the studies included in the review and in our own cohort. Given the consequences of detection of an underlying malformation, DSA should probably be considered, in particular in younger patients, unless there are distinct contra-indications. Based on the available data, presence or absence of hypertension or use of oral anticoagulation should not be a criterion, but the lack of association is probably a result of indication bias. In our small case series, we could not find a significant association between SVD and presence of macrovascular abnormalities, while this was found in a large prospective study among patients with non-traumatic ICH.²⁰ This study showed a higher probability of finding a macrovascular cause among patients without signs of SVD. The absence of an association in our study might be explained by a lack of power.

CTA and MRA may be preferred imaging modalities due to their non-invasive character but their diagnostic accuracy in isolated IVH is unsettled. A recent Cochrane review investigated the yield of non-invasive imaging in patients with spontaneous parenchymal intracerebral hemorrhage.²¹ Eleven studies, totaling 927 patients were identified that compared either CTA or MRA with DSA. Only five patients had isolated IVH. This meta-analysis showed a high sensitivity (95%; 95% CI 90–97%) and specificity (99%; 95% CI 97–100%) of CTA, and similar numbers for MRA, but results were limited by methodological drawbacks in some of the included studies. Given the small number of patients with isolated IVH included in this meta-analysis, generalizability to patients with isolated IVH is uncertain. A recent prospective cohort study of 300 patients with ICH (not including patients with isolated IVH) showed a lower sensitivity of CTA than found in the meta-analysis (74%; 95% CI 62–84%) and also a lower specificity (91%; 95% CI 86–94%) for detection of a macrovascular lesion as the underlying cause of ICH.²⁰

Several limitations of our study need to be addressed. First, most studies had small sample sizes and a retrospective design. Second, case finding methods varied among included studies, as well as selection criteria for angiography performance. This decreases comparability of the results and may partly explain the wide variation of yield of DSA. Third, information on age, history of hypertension and use of oral anticoagulants was only available at summary level, and no distinction was made between patients who did or did not undergo angiography. Also, some studies presented no information on risk factors and had to be excluded from these analysis. Fourth, not all patients included in our case series underwent angiographic imaging (CTA, MRA or DSA) nor was DSA performed in all patients with negative CTA. As a result, small vascular abnormalities may have been missed. A strength of our study is that we have summarized the available evidence in a systematic way with pre-specified inclusion criteria and without applying any language restrictions. Also, to the best of our knowledge, our case series of diagnostic yield in isolated IVH is the largest so far reported.

The results of this study suggest that the yield of angiographic examination is higher in younger patients with isolated IVH. The available data are insufficient to provide firm recommendations regarding in which patients what type of angiographic examinations should be performed, or in which patients angiographic examinations are not needed. Data on the yield of CTA and MRA in patients with isolated IVH are urgently needed.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: CJJ van Asch was supported by a Dutch Heart Foundation grant (grant no. 2007B048, to CJM Klijn). CJM Klijn is supported by a clinical established investigator grant of the Dutch Heart Foundation (grant no. 2012 T077), and an Aspasia grant from The Netherlands Organisation for Health Research and Development, ZonMw (grant no. 015008048).

Ethical approval

The Medical Ethics Review Committee of the University Medical Center Utrecht reviewed the protocol and waived the requirement for ethical review.

Informed consent

N/A

Guarantor

CJMK.

Contributorship

CJMK initiated the study. NAH and CJJvA performed the literature search. NAH screened title and abstract for eligible studies. Uncertainties about eligibility were resolved by discussion with CJJvA and CJMK. NAH and CJJvA extracted data from identified articles. NAH performed the statistical analyses. All authors interpreted the data. NAH and CJJvA drafted the manuscript, which was critically reviewed and revised for important intellectual content by all authors.

Acknowledgement

Ryan W. Snider and dr. Christine A.C. Wijman kindly provided additional data of their study on the diagnostic yield of CTA in addition to MRI/MRA in spontaneous intracerebral hemorrhage.

Note

References

Passero

Ulivelli

Reale

. Primary intraventricular haemorrhage in adults. Acta Neurol Scand 2002; 105: 115–119.

Roos

Hasan

Vermeulen

. Outcome in patients with large intraventricular haemorrhages: a volumetric study. J Neurol Neurosurg Psychiatry 1995; 58: 622–624.

Srivastava

Sannegowda

Satija

. Primary intraventricular hemorrhage: clinical features, risk factors, etiology, and yield of diagnostic cerebral angiography. Neurol India 2014; 62: 144–148.

Giray

Sen

Sarica

. Spontaneous primary intraventricular hemorrhage in adults: clinical data, etiology and outcome. Turk Neurosurg 2009; 19: 338–344.

Whiting

Rutjes

Reitsma

. The development of QUADAS: a tool for the quality assessment of studies of diagnostic accuracy included in systematic reviews. BMC Med Res Methodol 2003; 3: 25.

Bossuyt

Reitsma

Bruns

. Towards complete and accurate reporting of studies of diagnostic accuracy: the STARD initiative. Fam Pract 2004; 21: 4–10.

Snider

Thai

Narayana

. Diagnostic yield of CT angiography in addition to MRI/MRA in spontaneous intracerebral hemorrhage. Stroke 2010; 41: e310.

Abu Bakar

Shuaib

Mohd Ariff

. Diagnostic cerebral angiography in spontaneous intracranial haemorrhage: a guide for developing countries. Asian J Surg 2005; 28: 1–6.

Flint

Roebken

Singh

. Primary intraventricular hemorrhage: yield of diagnostic angiography and clinical outcome. Neurocrit Care 2008; 8: 330–336.

10.

Halpin

Britton

Byrne

. Prospective evaluation of cerebral angiography and computed tomography in cerebral haematoma. J Neurol Neurosurg Psychiatry 1994; 57: 1180–1186.

11.

Hameed

Khealani

Mozzafar

. Prognostic indicators in patients with primary intraventricular haemorrhage. J Pak Med Assoc 2005; 55: 315–317.

12.

Jayakumar

Taly

Bhavani

. Prognosis in solitary intraventricular haemorrhage. Clinical and computed tomographic observations. Acta Neurol Scand 1989; 80: 1–5.

13.

Marti-Fabregas

Piles

Guardia

. Spontaneous primary intraventricular hemorrhage: clinical data, etiology and outcome. J Neurol 1999; 246: 287–291.

14.

Toffol

Biller

Adams

Jr . The predicted value of arteriography in nontraumatic intracerebral hemorrhage. Stroke 1986; 17: 881–883.

15.

Weisberg

Elliott

Shamsnia

. Intraventricular hemorrhage in adults: clinical-computed tomographic correlations. Comput Med Imaging Graph 1991; 15: 43–51.

16.

Zhu

Chan

Poon

. 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.

17.

Darby

Donnan

Saling

. Primary intraventricular hemorrhage: clinical and neuropsychological findings in a prospective stroke series. Neurology 1988; 38: 68–75.

18.

Lin

Howng

. Nontraumatic intracerebral hemorrhage in young adult. Kaohsiung J Med Sci 1997; 13: 237–242.

19.

Yeung

Ahmad

Aviv

. Comparison of CTA to DSA in determining the etiology of spontaneous ICH. Can J Neurol Sci 2009; 36: 176–180.

20.

van Asch

Velthuis

Rinkel

. Diagnostic yield and accuracy of CT angiography, MR angiography, and digital subtraction angiography for detection of macrovascular causes of intracerebral haemorrhage: prospective, multicentre cohort study. BMJ 2015; 351: h5762.

21.

Josephson

White

Krishan

. Computed tomography angiography or magnetic resonance angiography for detection of intracranial vascular malformations in patients with intracerebral haemorrhage. Cochrane Database Syst Rev 2014; 9: CD009372.