Reversible cerebral vasoconstriction syndrome complicated by posterior reversible encephalopathy syndrome in a child triggered by systemic immunosuppression

RCVS and PRES are both infrequent neurological disorders, especially in children. RCVS is a clinical and radiological entity characterized by hyperacute onset of reversible, multifocal narrowing of the intracranial arteries, classically associated with thunderclap headache, hemorrhage, infarction, cytotoxic edema, and focal neurological deficits. While frequently idiopathic, RCVS can be provoked by vasoconstrictive drugs, pregnancy, and systemic illnesses.² PRES is classically associated with parieto-occipital vasogenic edema and headache, encephalopathy, visual disturbances, and seizures. The moniker may be deceptive, in that PRES is not always posteriorly predominant and as many as 20% experience long-term neurological sequelae.³ “Atypical” PRES with a nonclassical distribution of brain abnormalities often involving the basal ganglia, frontal lobes, and brainstem is more common in children; hemorrhage and infarction may be seen in severe cases.⁶ Uncontrolled hypertension, pre-eclampsia and eclampsia, cytotoxic agents, and renal dysfunction have all been implicated in the development of PRES.³

Shared radiological features of PRES and RCVS, similar inciting etiologies, and overlapping clinical features have led some to hypothesize a common pathophysiology, or even that both syndromes occur along a spectrum.¹ (Table 3) Coexistence of PRES occurs in 17-38% of those with RCVS.^2,7 Both syndromes have an acute, self-limited, and usually monophasic course involving headaches, encephalopathy, seizures, and focal neurological deficits. Both are associated with pregnancy, especially eclampsia, as well as hypertension, immunosuppressants, autoimmune syndromes, sepsis, and intravenous immunoglobulin.² There is a paucity of documented cases of pediatric RCVS, and even fewer recognized cases with overlapping features of PRES and RCVS in children (Table 4).^8-15 The clinical and imaging features may also overlap with other disorders such as central nervous system vasculitis.

PRES is thought to be a disorder of small-sizes arterioles, while RCVS is characterized by medium- and large-sized arterial dysfunction.⁸ Some have hypothesized that RCVS progresses centripetally from small, distal vessels to proximal medium- and large-sized arteries. Thus, cerebral angiography is often normal in the early stages in the disease, and helps explain why intracranial hemorrhage is typically seen in the first week and ischemia in the second week as vasoconstriction propagates.²

Chen and Wang¹⁶ recently proposed a model of the pathophysiology of RCVS, in which both predisposing and precipitating factors lead to endothelial dysfunction, sympathetic overactivity, and oxidative stress, resulting in disruption of the blood-brain barrier and failure of cerebrovascular autoregulation. Endothelial dysfunction is proposed as the common pathophysiologic mechanism for both RCVS and PRES. Endothelial dysfunction occurs in PRES due to either endogenous or exogenous cytokine release, mitigated by endothelial activation, vascular permeability, and vasogenic edema.³ Vasodilation in response to hypercapnia, an endothelium-dependent process, is impaired in RCVS, and highlights the impaired cerebrovascular tone in this condition.¹⁷ Elevated levels of catecholamines may elicit both vasoconstriction and vasodilation within the cerebral arteries of the same individual due to varying distribution of vasoactive receptors. Chen, et al., found that circulating microRNAs related to cerebral vascular tone and endothelial function, such as endothelin-1 (EDN1) and genes involved in the transforming growth factor-beta (TGF-β) could help differentiate those with RCVS from healthy controls.¹⁸ No definitive studies elucidating a common pathophysiological mechanism for RCVS and PRES exist, and classically, isolated RCVS occurs in the absence of vasogenic cerebral edema and isolated PRES occurs in the absence of cerebral vasoconstriction. Whether the two conditions are on a continuum within the same syndrome remains speculative.

There are no specific disease-modifying therapies for PRES. Treatment focuses on identifying and removing inciting triggers, and symptomatic management including addressing hypertension and dramatic fluctuations in blood pressure, anti-seizure medications as necessary, and prompt delivery of the fetus if secondary to preeclampsia or eclampsia.³ In RCVS, similar principles apply, including removal of exacerbating conditions, hemodynamic stability, and monitoring for secondary complications. Calcium channel blockers are recommended to promote vasodilation. These have been shown to improve headache severity, yet evidence that they change radiologic or clinical outcomes after RCVS is lacking.² Transcranial Doppler (TCD) may be useful for monitoring these patients, with several studies suggesting that TCD velocities increase in those with RCVS.^19-21 Avoidance of glucocorticoids, triptans, noradrenergic, serotoninergic, and other exacerbating medications is recommended.

Glucocorticoids, as well as tacrolimus and mycophenolate, were administered to our patient immediately prior to symptom onset. PRES has been linked to glucocorticoid administration in both adults and children, especially when coupled with concurrent cancer, active chemotherapy, or active immunosuppression, as in our patient.³ Steroid administration in RCVS is associated with poorer radiologic and clinical outcomes.² Tacrolimus, even in the absence of PRES, can produce insomnia, tremors, and encephalopathy, and PRES attributed to tacrolimus can occur at therapeutic levels;²² whether tacrolimus toxicity constitutes a separate entity or is also part of this spectrum is another consideration. Avoiding triggering medications is challenging when ongoing immunosuppression is required in patients with organ transplants, and transitioning from systemic immunosuppression can be challenging.

Our admittedly limited data suggest a male predominance for concurrent RCVS and PRES in children – seven of eight cases. A review of available cases of pediatric RCVS found a majority (65.5%) occurred in males,²³ in contrast to the adult population, where the majority of RCVS cases occur in females.² Both Thavamani, et al., and Raj, et al., found that pediatric PRES had a female predominance, of 56.5% and 60%, respectively,^4,5 which is congruent with adult data.²⁴

Two of the described cases of dual pediatric PRES and RCVS occurred in individuals with sickle cell disease (SCD) who subsequently received blood transfusion. Individuals with SCD are thought to be at increased risk of PRES,²⁵ in part due to known endothelial dysfunction in this population. Chronic anemia leads to hypoxia, and hemolysis will deplete levels of nitric oxide, both of which contribute to endothelial dysfunction.^26,27 In addition, leukocyte activation and platelet dysfunction seen in these individuals leads to increased levels of proinflammatory cytokines, similarly increasing endothelial dysfunction.²⁸ Sudden increases in hemoglobin are hypothesized to decrease physiologic cerebrovascular vasodilation, thereby increasing vascular tone and inducing nitrogen oxide scavenging, both may produce endothelial injury and subsequent vasogenic edema.^29,30 While definitive data are lacking, individuals with SCD may be at increased risk of developing PRES and RCVS, presumably during crises when hemolysis and transfusions occur.

A prior case of RCVS complicated by PRES was documented in a boy with Loeys-Dietz syndrome,¹⁰ an autosomal dominant connective tissue disorder due to mutation in TGFBR2, which occurred two months after aortic replacement following aortic root dilatation and aortic dissection. Akazawa, et al.,¹⁰ hypothesized that aberrant endothelial TGF-β signaling, which has been demonstrated in Loeys-Dietz syndrome,³¹ may have contributed. Our patient had previously undergone genetic testing for connective tissue disease, due to joint hyperlaxity and congenital heart disease. Testing revealed two heterozygous variants of uncertain significance in LZTS1 (c.1028G>A; p.Arg343Gln), the leucine-zipper, putative tumor suppressor-1, which has been associated with autosomal dominant forms of hypermobile Ehlers-Danlos syndrome, but whether it constitutes a form of vascular Ehlers-Danlos syndrome is unclear.³² Further studies on impairment in cerebral autoregulation and endothelial dysfunction in those with connective tissue disorders are warranted.

Obtaining cerebrovascular imaging for patients with PRES to assess for comorbid RCVS has been suggested but is not widely considered standard of practice. While MRA has been validated for evaluation for RCVS in adults,³³ its validation was based on proximal branches of intracranial arteries and may miss more subtle, distal irregularities, such as those seen in our patient. Performing MRA through the whole head to include the distal arteries may increase the yield. Whether cerebrovascular imaging should be performed, and whether MRA or CTA is the more optimal modality, in children with PRES and concern for concurrent RCVS has yet to be fully elucidated.