Hyperintense vessels: An early MRI marker of reversible cerebral vasoconstriction syndrome?

Reversible cerebral vasoconstriction syndrome (RCVS) is an increasingly recognized clinical emergency characterized by acute onset of severe headaches (usually multiple thunderclap headaches) and reversible segmental stenosis of cerebral arteries (1). RCVS is associated with several potentially devastating complications or comorbidities, such as posterior reversible encephalopathy syndrome (PRES) (9%–38%), ischemic stroke (4%–54%), intracerebral hemorrhage (12%–20%), cortical subarachnoid hemorrhage (SAH) (22%–34%), or cervical artery dissection (1%–12%) and even death (2–7). The prevalence of RCVS is unknown, but it could be higher than that of its most important differential diagnosis—aneurismal SAH when patients complain of thunderclap headaches.

Although the clinical features of RCVS are distinct, the diagnosis could sometimes be difficult in the lack of awareness or alertness about the dynamic nature of the disease. This is especially the case in the very early stage of the disease, when cerebral vasoconstrictions might be insignificant on the initial angiographic studies (3,4), and ischemic or hemorrhagic complications might not have occurred. In this issue of Cephalalgia, Kameda et al. report on a case of a 30-year-old woman who developed thunderclap headaches and seizure one week post-partum (8). Magnetic resonance imaging disclosed hyperintense vessel signs and PRES on the initial fluid-attenuated inversion recovery (FLAIR) imaging in the absence of significant vasoconstriction on magnetic resonance angiography (MRA). The hyperintense vessels disappeared one week later, but vasoconstrictions of major cerebral arteries became apparent (8). Although this case is not the first one to report hyperintense vessel sign in patients with RCVS (9,10), the unique finding that hyperintense vessels predated the occurrence of cerebral vasoconstrictions raised several important questions: 1.) Could hyperintense vessel sign help diagnose RCVS in the very early stage of disease? 2.) Does this finding support the centripetal propagation theory? and 3.) How to differentiate hyperintense vessels from cortical SAH, the early complication of RCVS?

Hyperintense vessels on FLAIR imaging are hypothesized to be distal branches of the cerebral arteries or their leptomeningeal anastomotic collateral vessels with very low-flow velocities attributed to severe stenosis or occlusion of proximal cerebral arteries (11). In RCVS, FLAIR hyperintense vessels could be noted in up to 22% of patients (10). The presence of hyperintense vessels in RCVS is associated with more severe vasoconstrictions and cerebral hemodynamic derangements, and could herald the occurrence of ischemic stroke or PRES (10). In the case reported by Kameda et al. (8), hyperintense vessels were noted concomitantly with PRES two days after the onset of thunderclap headaches, in the absence of MRA vasoconstrictions. If these FLAIR hyperintense vessels also came from the failure of cerebrospinal fluid (CSF) nulling effect due to stagnant or retrograde flow of distal arterioles or leptomeningeal networks, this rare observation might have serendipitously captured the point when severe vasoconstrictions were still restricted to distal arteries beyond the resolution of routine MRA. In our previous study, MRA vasoconstrictions were not at its worse point until 11–20 days after headache onset (4). In another large series, it was reported that up to 33% of the patients with RCVS had a normal early angiogram (3). In other words, in one-third of the patients, the diagnosis of RCVS could be missed if their initial MRA were performed within 10 days of headache onset. In these patients, identifying hyperintense vessels on FLAIR could be helpful for an early diagnosis before a second scan. However, without a large-scale prospective study, we still don’t know if this finding could serve as a generalizable objective early imaging sign for RCVS or just highly exceptional. In fact, another possibility is that these hyperintense vessels could be an accompanying imaging sign of PRES. In our study, eight out of the 95 RCVS patients had PRES, and all of them had hyperintense vessels noted on the same scan of PRES, although all our patients had significant MRA vasoconstrictions concomitantly (10).

Based on the timing and possible pathophysiology of the early complications PRES and cortical SAH, Ducros et al. proposed a plausible mechanism for the temporal evolution of the vasoconstrictions: The arterial abnormalities first involve small distal arteries and then progress toward medium- and large-sized vessels (3,5). The imaging finding of this case (8) is good evidence in support of this centripetal propagation theory. In our previous study, we hypothesized that these hyperintense and perhaps more engorged vessels (i.e. distal vasodilatation, either compensatory or de novo) might facilitate the activation of perivascular nociceptors and contribute to the anatomical distribution of pain (10). This might explain the disconcordant temporal evolution of thunderclap headaches and vasoconstriction (4), which was also true in this case (8).

Because both hyperintense vessels and cortical SAH can be noted during the early stage of the disease in at least 20% of patients (5,10), and both appear hyperintense on FLAIR imaging, it is mandatory to differentiate the two. It’s usually not very difficult because distal hyperintense vessels usually appear as well-circumscribed dot- or tubular-like hyperintense lesions running through the dark CSF space, whereas the hyperintensities of cortical SAH are usually diffuse and occupy the entire sulcus. However, when the blood amount of cortical SAH is very few, or in the presence of motion artifact, T2* gradient echo or susceptibility-weighted imaging would be helpful for diagnosis. On the other hand, whether the imaging findings are hyperintense vessels or cortical SAH, even in the absence of conspicuous vasoconstriction, a diagnosis of RCVS should always be considered when seeing a patient with multiple thunderclap headaches.

Although much remains unknown, it is exciting to see that the publications about RCVS have soared in the past few years, rapidly enhancing our recognition of this disorder. Each single case is an important piece of the jigsaw puzzle; by putting them together, we will see the whole picture.