Reversible cerebral vasoconstriction syndrome: Is it more than just cerebral vasoconstriction?

Abstract

Background

Systemic vascular alterations have not been described in reversible cerebral vasoconstriction syndrome (RCVS). We present a case series of RCVS patients having cardiac dysfunction during ictus, with a subset showing complete resolution of cardiomyopathy.

Methods

Retrospective case-series: Cardiac left ventricular ejection fraction (LVEF) and wall motion abnormalities (WMA) visualized on transthoracic echocardiography (TTE), performed during RCVS ictus and follow-up was analyzed.

Results

Of 68 patients, 18 (26%) had a TTE performed around ictus. Three of 18 (17%) patients demonstrated WMA on initial TTE. All three patients were female without previous coronary artery disease or heart failure, and were asymptomatic from the cardiac dysfunction. WMA resolved completely on follow-up in Patients 1 and 2. Global LV dysfunction persisted for at least 90 days in Patient 3.

Conclusion

Although the exact pathophysiology of the cardiomyopathy is uncertain, it may be related to localized coronary vasoconstriction causing myocardial ischemia/infarction. Vasoconstriction may not be limited to the cerebral vasculature and may involve extracerebral organs. Cardiac ventricular abnormalities may be a part of the RCVS spectrum.

Introduction

Reversible cerebral vasoconstriction syndrome (RCVS) is characterized by thunderclap headache, with or without neurological deficits, with evidence of reversible cerebral vasoconstriction on cerebral angiography (1). The pathophysiology of RCVS is thought to be secondary to alteration of intracerebral vascular tone. Systemic vascular alterations have not been described. We present a case series of RCVS patients having cardiac dysfunction during ictus, with a subset showing complete resolution of cardiomyopathy. We hypothesized that vasoconstriction in RCVS may not be limited to the cerebral vasculature.

Methods

A retrospective analysis of patients diagnosed with probable or definite RCVS at our institution from 1990 to 2013 was conducted. Medical records were reviewed for demographics, symptoms and cerebral angiography. We analyzed cardiac left ventricular ejection fraction (LVEF) and presence of wall motion abnormalities (WMA) visualized on transthoracic echocardiography (TTE), performed during the ictus of RCVS and at follow-up if available. A cardiologist independently reviewed all abnormal TTEs and electrocardiograms (EKGs).

Results

Sixty-eight patients were included in the study. A TTE performed around the ictal stage was available in 18 (26%) patients including 16 (89%) women with mean age of 43 ( ± 12 SD) years. TTE was performed for stroke etiology evaluation in 17/18 patients. One patient also had chest pain at ictus for which TTE was recommended. Three of 18 (17%) patients demonstrated echocardiographic WMA on initial screen (Table 1). They were all female, and two were postpartum. All three patients had no previous history of coronary artery disease (CAD) or congestive heart failure (CHF). Patients 1 and 3 were asymptomatic from the cardiac dysfunction during the ictal stage and lacked evidence of acute cardiac ischemia on serial electrocardiograms or elevation of cardiac enzymes. Patient 2 was also asymptomatic, but had mild isolated elevation of creatinine kinase with non-specific EKG changes, insufficient to categorize as myocardial infarction (MI). Cardiac dysfunction improved on repeat TTE in Patients 1 and 2, two and seven days after initial TTE, respectively. Most remarkable was the complete resolution of WMA and hypokinesis. Global left ventricular (LV) dysfunction persisted for at least 90 days in Patient 3. All patients had diffuse vasoconstriction on initial cerebrovascular imaging and its resolution in Patients 2 and 3 started five and 28 days after symptom onset, respectively, with eventual complete normalization. Reversal of vasoconstriction was not documented in Patient 1 since she was doing well and declined further testing. Of the remaining 15 patients without echocardiographic abnormalities, none had a previous history of CHF or cardiac symptoms around the ictus. Only one of 15 patients had transient elevation of cardiac enzymes and inferolateral ST-depression on EKG consistent with Type 2 non-ST elevation MI.

Table 1.

Clinical characteristics, echocardiographic and cerebrovascular imaging findings.

Pt.	Age/sex	Trigger	Symptoms	Initial brain imaging	Initial TTE/days post-symptom onset	Follow-up TTE/days post-symptom onset	Days post-symptom onset when cerebral vasoconstriction normalization started
1	25/F	Gravida 3, eclampsia	Thunderclap headache post-delivery day 1	ICH post-right temporal lobe Diffuse vasoconstriction	Day 3 LVEF ∼50% Inferior wall hypokinesis	Day 5 LVEF ∼55% No WMA/hypokinesis	Repeat vessel imaging not performed
2	57/F	Head trauma	Dizziness	Ischemic stroke Bi-parietal, left occipital and left front lobes Diffuse vasoconstriction	Day 5 LVEF ∼45% Hypokinesis in left anterior descending artery territory	Day 12 LVEF ∼60% No WMA/hypokinesis	Day 5
3	36/F	Gravida 4, uncomplicated delivery	Acute headache post-delivery day 0, acute-onset hemiparesis three weeks post-delivery	Ischemic stroke Right parietal and front lobe Diffuse vasoconstriction	Day 24 LVEF ∼40% Global hypokinesis	Day 90 LVEF ∼40% Global hypokinesis	Day 28

F: female; ICH: intracerebral hemorrhage; TTE: transthoracic echocardiography; LVEF: left ventricular ejection fraction; WMA: wall motion abnormalities.

Discussion

We present three patients without previous CAD who developed cardiomyopathy, most likely during the ictus of RCVS. In two patients, the echocardiographic abnormalities were transient, and resolved completely much like the reversal of cerebral vasoconstriction. Although test-retest variability could account for changes in the LVEF, the resolution of WMA on comparison was striking. The improvement in cardiac function occurred within five and 12 days of symptom onset, closely paralleling cerebral vasoconstriction normalization. At last follow-up three months from symptom onset, global hypokinesis persisted in Patient 3. Although we hypothesize that onset of cardiomyopathy was concurrent with the ictus of RCVS, the cardiac disease could be unrelated to RCVS. Other potential differential diagnoses include peripartum cardiomyopathy or cardiomyopathy secondary to hypertensive disease in pregnancy as explained below.

The heart-brain connection causing acute stress-induced LV dysfunction (2) has been described with Takotsubo cardiomyopathy (3), and after subarachnoid hemorrhage causing neurocardiogenic stunning (4). With Takutsubo, transient LV WMA occurs in the apical and/or midventricular segments, with extension beyond a single epicardial coronary artery distribution, along with electrocardiographic changes. This was not the pattern seen in our patients. Another possibility specific to pregnancy is peripartum cardiomyopathy (5) and cardiomyopathy related to hypertensive pregnancy disorders (6), the etiology of which remains incompletely understood. The former has a variable course, with some showing rapid clinical and echocardiographic recovery, while others have persistent long-term cardiac dysfunction. In stress-induced cardiomyopathy, coronary spasm and myocardial adrenergic stimulation related to catecholamine excess from autonomic storming is postulated as the mechanism (2). Although we are uncertain of the exact pathophysiology of the cardiomyopathy in our patients, it may be related to localized coronary vasoconstriction causing myocardial ischemia/infarction. Indeed RCVS may be related to stress-induced, peripartum and pre-eclampsia/eclampia-induced cardiomyopathy and could represent a spectrum of the same pathophysiologic process in different vascular beds.

Extracranial vessel involvement in RCVS has been reported with occurrence of cervical artery dissection. The association between these two conditions was observed in 12% of patients with RCVS, and 7% of patients with cervical artery dissection in a prospective series (7). It is difficult to elucidate whether RCVS or dissection starts first, and whether RCVS is triggered by cervical artery dissection. Extracerebral vascular abnormalities including cardiac dysfunction have not been described in RCVS. Field et al. described a case of RCVS associated with internal carotid artery dissection, and a tight, non-atherosclerotic narrowing of the right renal artery at its mid-to-distal segment (8). However, the authors suggest that a unifying diagnosis would be fibromuscular dysplasia affecting both the carotid and renal arteries, and cerebral vasoconstriction developed after the carotid dissection.

This study has multiple limitations. Our cohort does not represent consecutive cases, and in the wide clinical spectrum of RCVS likely represents those with severe vasoconstriction with ischemic or hemorrhagic strokes. Echocardiograms were unavailable for many patients, either because of lack of medical records or because they were not performed around the ictus of RVCS. For the three patients with echocardiographic abnormalities, we do not have prior echocardiograms to compare, but the quick normalization of cardiac dysfunction suggests that it is likely related to the underlying vasoconstrictive pathophysiology of RCVS.

In conclusion, vasoconstriction may not be limited to the cerebral vasculature in RCVS and may involve extracerebral organ systems. Cardiac ventricular function abnormalities may be a part of the spectrum of RCVS. Our study is hypothesis generating, and further prospective studies of RCVS should investigate this hypothesis.

Clinical implications

The pathophysiology of reversible cerebral vasoconstriction syndrome (RCVS) is thought to be secondary to alteration of intracerebral vascular tone. Systemic vascular alterations have not been described.

We describe a case series of RCVS patients who developed cardiac dysfunction during the RCVS ictus. Three of 18 (17%) patients developed left ventricular wall motion abnormalities (WMA) seen on transthoracic echocardiography performed during RCVS ictus. In two of these patients, there was complete resolution of WMA within two weeks.

Although the exact pathophysiology of the cardiomyopathy is uncertain, it may be related to localized coronary vasoconstriction causing myocardial ischemia/infarction.

Vasoconstriction may not be limited to the cerebral vasculature in RCVS and may involve extracerebral organs. Cardiac ventricular abnormalities may be a part of the RCVS spectrum.