When do you recommend MRA in patients with spontaneous ICH and negative CTA?

1. In short, never. MRA, specifically, is not superior to CTA for determining if there is abnormal vascular anatomy as a cause of a spontaneous ICH.

2. However, the use of MR imaging can be very useful. In general, I recommend waiting 8-12 weeks to allow the blood to clear so that an underlying lesion may be seen. However, if there are other reasons to believe that there may be an underlying mass (e.g., a known primary cancer), then MR with Gd contrast earlier can be useful. You may be able to uncover a tumour, AVM or AVF with this kind of imaging.

3. High-resolution T1-fat.sat images through the large arteries of interest ("vessel wall imaging") may be helpful if you suspect that the underlying problem might be a large artery vasculopathy ...

I would add that the new AHA ICH guidelines include more specific criteria for when to consider angiogram even if CTA is negative, including in those younger than 45 as well as those 45-70 with no history of hypertension for example. Read more here: Greenberg et al., PMID 35579034.

Usually, if there is a good quality CTA, there is no need for repeat non-invasive vascular imaging like MRA.

If the patient is young and the bleeding is not classical for hypertensive bleed, we start with an MRI brain with and without contrast in the acute setting as well as repeat it later on after 6-8 weeks to give time for resolution of blood and to see if there is an underlying cavernoma (which will be occult for any vascular imaging including DSA), or tumor.

On some occasions, we consider DSA as a vessel imaging to identify blister-like aneurysms or dural AVF which will not be likely occult in non-invasive imaging.

MS