Thu Le

Greg Zaharchuk

We chose to study this topic after we started routine implementation of arterial spin-labeling (ASL) imaging on our clinical magnets.  We noticed the obvious findings from large AVMs, but were surprised to see how well more subtle, smaller AVMs and dural AVFs showed up on ASL.  This is often a difficult diagnosis to make with any imaging modality besides digital subtraction angiography, and we realized that it might be possible to identify these lesions more accurately with ASL MRI.

We believe that the presence of venous signal on ASL is concerning for a shunt-type lesion; while it can be seen in some other specific situations (seizure, subacute stroke, and in children with high flow), in our experience, it is most often associated with either an AVM or dural AVF.  This lesion may be quite small and not appreciated on other imaging sequences, particularly if there is intracranial hemorrhage.  For this reason, one should consider a shunt in the differential diagnosis and consider whether DSA might be clinically useful for further work-up and treatment.  In patients who have equivocal findings for a vascular malformation, and who might be poor candidates for DSA, a negative ASL study may tip the balance towards not performing the angiogram.

We often receive referrals for “rule out dAVF or small AVM” from clinicians who want more information before deciding on whether to refer patients on to the more invasive cerebral angiogram study.  In fact, ASL-negative shunt lesions are so uncommon that our angiographers often decide to forgo DSA in the setting of a good quality negative ASL study.  On the other side, we have seen patients in whom vascular malformations were not clinically suspected, but in whom the ASL study suggests such a diagnosis, who go on to have DSA for confirmation.

We have received feedback from others who have used ASL for this clinical indication.  One of our former fellows sent us a case of a patient with headaches who had gone undiagnosed for 11 years, with other misdiagnoses being raised, including MS and stroke.  ASL demonstrated obvious high signal in the sphenoparietal vein, diagnostic of a shunt lesion.

We would like to perform a prospective study to extend these findings, because the initial study was retrospective and could be subject to unknown biases.  We also plan write up a case series of all patients in our institution who had a final diagnosis of AVM or dural AVF, and who underwent ASL imaging.  Ultimately, we would like to do a cost-benefit analysis of the use of ASL MRI to determine the need for invasive studies such as DSA.

We have written a case report of a particularly interesting patient who presented with unilateral tinnitus who underwent an essentially negative CTA and conventional MRI study, but whose ASL demonstrated a focal abnormality of the marginal sinus, which was confirmed with angiography as a Cognard class I dural AVF.¹ The patient deferred treatment and returned 7 months later, at which time his tinnitus had stopped;  the abnormal ASL signal vanished, and DSA confirmed spontaneous thrombosis of the dural AVF.

We also have a manuscript under consideration examining the ASL and bolus perfusion-weighted imaging findings in a large group of developmental venous anomalies, about 8% of which show shunting on ASL imaging;  we think these may represent transitional lesions with AV shunting.

Thu Le is now a radiologist at Visalia Imaging, Visalia, CA.

References

1. Alexander M, McTaggart R, Santarelli J, et al. Multimodality evaluation of dural arteriovenous fistula with CT angiography, MR with arterial spin labeling, and digital subtraction angiography: case report. J Neuroimaging, published online June 7, 2013. doi: 10.1111/jon.12032

 

Read this article at AJNR.org . . .