High sensitivity and specificity of 4D-CTA in the detection of cranial arteriovenous shunts

Matthijs In ’T Veld,Marlise P Dos Santos,Frederick J A Meijer,Rolf Fronczek,Peter W A Willems,Marianne A A Van Walderveen

doi:10.1007/s00330-019-06234-4

Abstract

PurposeIn a prospective cohort study, we evaluated the diagnostic accuracy of time-resolved CT angiography (4D-CTA) compared to digital subtraction angiography (DSA) for detecting cranial arteriovenous shunts.Material and methodsPatients were enrolled if a DSA had been ordered querying either a dural arteriovenous fistula (dAVF) or a cerebral arteriovenous malformation (bAVM). After enrolment, both a DSA and a 4D-CTA were performed. Both studies were evaluated using a standardized form. If a dAVF or bAVM was found, its classification, angioarchitectural details, and treatment options were recorded.ResultsNinety-eight patients were enrolled and 76 full datasets were acquired. DSA demonstrated a shunting lesion in 28 out of 76 cases (prevalence 37%). 4D-CTA demonstrated all but two of these lesions (sensitivity of 93%) and produced one false positive (specificity of 98%). These numbers yielded a positive predictive value (PPV) of 96% and a negative predictive value (NPV) of 96%. Significant doubt regarding the 4D-CTA diagnosis was reported in 6.6% of all cases and both false-negative 4D-CTA results were characterized by such doubt.Conclusions4D-CTA has very high sensitivity and specificity for the detection of intracranial arteriovenous shunts. Based on these results, 4D-CTA may replace DSA imaging as a first modality in the diagnostic workup in a large number of patients suspected of a cranial dAVF or bAVM, especially if there is no doubt regarding the 4D-CTA diagnosis.Key Points• 4D-CTA was shown to have a high diagnostic accuracy and is an appropriate, less invasive replacement for DSA as a diagnostic tool for cranial arteriovenous shunts in the majority of suspected cases.• Doubt regarding the 4D-CTA result should prompt additional DSA imaging, as it is associated with false negatives.• False-positive 4D-CTA results are rare, but do exist.

Highlights

Intracranial arteriovenous malformations (AVMs) represent abnormal connections between arteries and veins. They may be divided into brain AVMs and dural AVMs depending on whether the feeders are pial or dural arteries. brain (pial) AVMs (bAVMs) may be further divided into nidus-type AVMs, if there is a tangle of abnormal vessels, and arteriovenous fistulas (AVFs), if there is no intervening network [1, 2]
4D-computed tomographic angiography (CTA) misdiagnosed over 10% of the AV shunts (3 out of 28; 10.7%), in one of these, a shunt was found on 4DCTA but erroneously diagnosed as a bAVM instead of a dural arteriovenous fistula (dAVF)
4DCTA adds time-resolved cross-sectional imaging and visualizes non-vascular tissues, which may prove beneficial during surgery, especially as it may be used in frameless stereotaxy. We demonstrated these differences between 4D-CTA and digital subtraction angiography (DSA) in patients known to have a dAVF or bAVM [15,16,17,18]

Summary

Introduction

Intracranial arteriovenous malformations (AVMs) represent abnormal connections between arteries and veins. They may be divided into brain (pial) AVMs (bAVMs) and dural AVMs depending on whether the feeders are pial or dural arteries. DAVFs may drain solely through venous sinuses in which case they do not represent a risk of hemorrhage. BAVMs and dAVFs may be asymptomatic or present with a wide variety of symptoms such as focal neurological deficits, headache, bruit, or seizures. All of these lesions are characterized by early venous filling caused by the lack of passage of blood through a normal capillary network. But can result in transient or permanent neurological damage [10,11,12]

Objectives

Methods

Results

Discussion

Conclusion