Abstract
Background
The temporal and anatomical features of vasoconstriction in patients with reversible cerebral vasoconstriction syndrome within hours after symptom onset, in the hyperacute phase, are unclear.
Case result
Herein we report the cases of two patients with acute severe headache who were diagnosed with reversible cerebral vasoconstriction syndrome. Magnetic resonance imaging within hours after symptom onset revealed multiple areas of isolated cortical vasogenic edema and hyperintense vessel signs of the distal cerebral arteries. Follow-up imaging performed four days later in both cases showed diffuse segmental arterial vasoconstriction in the proximal regions of the cerebral arteries. Both patients received antivasoconstrictive therapy shortly after admission, and neither had neurological sequelae at discharge. The magnetic resonance imaging findings improved gradually within three months after symptom onset.
Conclusion
Isolated cortical vasogenic edema and hyperintense vessel signs, when observed within hours from sudden severe headache onset, may be useful early markers of reversible cerebral vasoconstriction syndrome.
Keywords
Introduction
Reversible cerebral vasoconstriction syndrome (RCVS) is characterized by the sudden onset of severe headache, a thunderclap headache, with reversible segmental vasoconstrictions on brain imaging that usually resolve within three months (1). It has been previously reported that RCVS may present with posterior reversible encephalopathy syndrome (PRES), hyperintense vessel signs (HVS), cerebral infarction, and cerebral hemorrhage, with each of these occurring in approximately 10–20% of patients (1,2,4). PRES is considered a vasogenic edema, as it shows hyperintensity on both fluid-attenuated inversion recovery (FLAIR) and the apparent diffusion coefficient (ADC) map, and HVS signifies slow blood flow due to cerebral vasoconstriction. However, the temporal and anatomical features of cerebral arterial vasoconstriction within hours after symptom onset, in the hyperacute phase of RCVS, remain unclear.
Herein we report the cases of two patients with acute onset of severe headache who were diagnosed with RCVS. Brain magnetic resonance imaging (MRI) performed within hours after symptom onset revealed multiple areas of isolated cortical vasogenic edema and HVS in branches of the cerebral arteries.
Case reports
Case 1
A 34-year-old hospitalized woman suddenly developed thunderclap headache. She had received a bone marrow transplant owing to acute lymphoblastic leukemia and had started therapy with an antineoplastic agent, etoposide (100 mg/day), four days before symptom onset. When first examined after thunderclap headache, neurological examination on admission yielded unremarkable findings. Brain computed tomography (CT) and cerebrospinal fluid examination conducted soon after the onset of headache yielded normal findings. Brain MRI performed nine hours after symptom onset showed multiple areas of isolated cortical vasogenic edema in regions supplied by both anterior cerebral arteries (ACAs) (Figure 1 (a)–(c)) and HVS in distal segments of both ACAs and middle cerebral arteries (MCAs) (Figure 1 (d)–(f)). Follow-up MRI four days after symptom onset revealed diffuse segmental arterial vasoconstriction in the proximal regions of the vertebrobasilar artery (Figure 1 (g)–(h)), whereas the isolated cortical vasogenic edema had resolved. She was treated with the calcium channel blocker verapamil, 120 mg/day, starting on day four. The headache alleviated on day 22, and she eventually recovered without any neurological complications. Brain MRA at 31 days was normal (Figure 1 (i)).
In Case 1, magnetic resonance imaging (MRI) was performed nine hours after symptom onset. (a) The diffusion-weighted imaging (DWI) was unremarkable. (b) The apparent diffusion coefficient (ADC) showed slightly high intensity in the cortex (arrow), and (c) the fluid attenuated inversion recovery (FLAIR) showed high intensity in the cortex (arrow), suggesting vasogenic edema. (d–f) Hyperintense vessel signs (HVS) were observed in regions of the middle cerebral artery in FLAIR (arrow). (g, h) Magnetic resonance angiography showed no remarkable change nine hours after symptom onset, but revealed diffuse segmental arterial vasoconstriction in the proximal region of the cerebral arteries after four days (arrow). (i) The cerebral arterial vasoconstrictions disappeared after 31 days.
Case 2
A 54-year-old woman with thunderclap headache and numbness on her left upper extremity was transferred to our emergency department. Neurological examination on admission revealed no deficits except for severe headache. Brain CT and cerebrospinal fluid examination yielded normal findings. Brain MRI five hours after symptom onset revealed multiple areas of isolated cortical vasogenic edema in regions of both ACAs (Figure 2 (a)–(c)). Follow-up MRI four days later revealed HVS in the regions of the MCAs (Figure 2 (d)) and diffuse segmental arterial vasoconstriction in the proximal cerebral arteries, including both internal cerebral arteries, left posterior cerebral artery, and right vertebral artery (Figure 2 (e-1), (e-2), (f-1), (f-2)), whereas the isolated cortical vasogenic edema had resolved. She was treated with the calcium channel blocker verapamil, 120 mg/day, on day one. According to the frequency of headache attack, we increased verapamil up to 240 mg/day and added intravenous magnesium sulfate 80mEq/day and continuous infusion of nicardipine 1–2 mg/hour. The headache disappeared on day 11 and she was discharged, without any neurological complications, on day 29. Brain MRA on day 52 revealed improvement of vasoconstriction (Figure 2 (g-1), (g-2)).
In Case 2, MRI was performed five hours after symptom onset. (a) The DWI was unremarkable. (b) The ADC showed high intensity in the cortex (arrow), and (c) the FLAIR also showed high intensity (arrow). (d) HVS were observed in regions of both middle cerebral arteries in FLAIR (arrow) four days after symptom onset. (e-1, e-2) There were no abnormal findings in MRA five hours after symptom onset, and (f-1, f-2) multiple segmental vasoconstrictions were observed in the proximal region of the cerebral artery four days after symptom onset (arrow). (g-1, g-2) Cerebral arterial vasoconstrictions were improved 52 days after symptom onset.
Discussion
It was noteworthy, in our cases, that isolated cortical vasogenic edema and HVS on FLAIR and the ADC map preceded the appearance of cerebral vasoconstriction on MRI. Further, the transient isolated cortical vasogenic edema was located in the distal areas of the cerebral anterior circulation on MRI in the hyperacute phase of RCVS.
Vasogenic edema shows high intensity in both FLAIR and ADC, and is known to be associated with blood-brain barrier (BBB) breakdown, which shares the pathology of PRES. A recent study using contrast-enhanced FLAIR images for 72 patients with thunderclap headache has shown that BBB breakdown was visualized in 23 of 41 RCVS patients and frequently located in regions of ACAs and MCAs including near the falx and convexity (5). PRES is found in approximately 10% of patients with RCVS within one week from onset (2), and mainly involves the regions of PCAs. Actually, in our cases, there is a possibility that PRES was complicated by RCVS. Given the findings of previous studies and our cases, however, isolated cortical vasogenic edemas in the distal regions of the cerebral anterior circulation on MRI suggest BBB breakdown in RCVS rather than PRES. In patients with sudden severe headache, MRI examination may be useful for early markers of RCVS.
HVS signifies decreased blood flow due to vascular stenosis, which is found in approximately 20% of patients with RCVS (4). The emergence of vasoconstriction on MRA tends to be delayed. In a previous report of 67 patients with RCVS, MRA abnormality was, in some cases, observed two weeks from onset (2). Moreover, there is a case report that showed cerebral vasoconstriction on MRA after HVS (3). From the time-course of these findings of HVS and vasogenic edema, it is hypothesized that arterial vasospasm first involves the distal branches and then progresses toward the proximal region of the cerebral arteries (2,5). Our two cases were consistent with this hypothesis.
Although the long-term prognosis of RCVS is generally favourable, clinical exacerbation occurs in approximately 34% of patients (6), and permanent disability or death in 5–10% (7). The timely diagnosis of RCVS is therefore important in clinical settings; however, it is often challenging owing to the limitations of MRA, which is often performed in suspected cases of RCVS for the evaluation of small, distal cerebral arteries (8). Hence, it is helpful for the early diagnosis of RCVS to focus on both MRA abnormalities and findings related to distal branch lesions, such as isolated cortical vasogenic edema and HVS seen on FLAIR and the ADC map. A recent study has suggested that BBB breakdown is an independent predictor of neurological complication in patients with RCVS (5) although our patients did not undergo gadolinium enhanced MRI. Because RCVS have characteristic image findings from the hyperacute phase, clinicians should perform MRI in patients with sudden severe headache.
In conclusion, we reported two cases of isolated cortical vasogenic edema and HVS with MRI during the hyperacute phase of RCVS. Although future studies are needed to validate our findings, MRI focusing on isolated cortical vasogenic edema and HVS in patients with severe headache may be a useful early marker to the diagnosis of RCVS.
Clinical implications
Isolated cortical vasogenic edema and hyperintense vessel sign (HVS), when observed within hours from sudden severe headache onset, may be useful early markers of reversible cerebral vasoconstriction syndrome (RCVS). MRI should be performed in patients with sudden severe headache because RCVS have characteristic imaging findings in the hyperacute phase.
Footnotes
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
