Childhood Central Nervous System Vasculitis

Fig. 40.1

Diagnostic evaluation of a child with suspected inflammatory brain disease. Abbreviations: CVID common variable immunodeficiency, HLH hemophagocytic lymphohistiocytosis, HES hypereosinophilic syndrome, vWF von Willebrand factor antigen, LAC lupus anticoagulans, ANA antinuclear antibody, dsDNA anti-double-stranded DNA antibody, ENA antibodies against extractable nuclear antigen, APL antiphospholipid antibody, TTG tissue transglutaminase antibody, TPO thyroid peroxidase antibody, anti-NMDAR N-methyl-D-aspartate receptor antibody, NMOIg neuromyelitis optica antibody also anti-aquaporin 4 antibody, anti-LGI leucine-rich glioma-inactivated 1 antibody, GAD glutamic acid decarboxylase antibody, TB tuberculosis

Primary CNS Vasculitis

Primary angiitis of the central nervous system (PACNS) is the most common inflammatory cause of severe, acquired neurological deficits in a previously healthy child [7]. Cravioto first described PACNS in adults in 1959 [10]. Initial case reports were almost exclusively based on autopsy data and revealed a granulomatous inflammation of the cerebral arteries. In 1988, Calabrese et al. described eight new cases and summarized the available literature of PACNS in adults [11]. The term PACNS was coined and diagnostic criteria for adults were proposed. These criteria mandate (1) a newly acquired neurological deficit, (2) angiographic and/or histological evidence of CNS vasculitis, and (3) the absence of a systemic condition that could explain these findings [11]. These Calabrese criteria were adopted and modified for childhood PACNS (cPACNS) requiring a newly acquired neurological deficit and/or psychiatric symptom in a patient ≤18 years of age [12]. The current classification of cPACNS is based on affected cerebral vessel size and disease course [12, 13]. CNS vasculitis in children is based on vascular disease. In adults clots and arteriosclerosis are often seen; this is not the case in children [7].

Childhood PACNS is divided in two subcategories: large vessel disease, angiography positive, and small vessel disease, angiography negative and brain biopsy positive [12, 13]. Three subtypes are recognized: (1) nonprogressive (NP) large-medium vessel cPACNS (angiography positive); (2) progressive (P), large-medium vessel cPACNS (angiography positive); and (3) small vessel (SV) cPACNS (angiography negative, biopsy positive). The three subtypes display distinct presenting symptoms, laboratory findings, disease course, and treatment outcome [12–14].

Angiography-Positive Large Vessel cPACNS

The term angiography-positive childhood primary angiitis of the central nervous system is used to describe a group of patients with involvement of the medium to large cerebral vessels. Angiography-positive cPACNS is further divided into two subtypes based on the course of the disease: NP large vessel cPACNS and P large vessel cPACNS [12]. NP disease is more common (68 %) [12]. In both subtypes boys are more commonly affected than girls. P-cPACNS is defined by progression on neuroimaging 3 months after initial angiography [12].

Angiography-Positive NP-cPACNS

Children with NP-cPACNS present with sudden-onset focal neurological deficits [12, 15]. These children are frequently diagnosed with arterial ischemic stroke. In childhood strokes more boys are affected than girls. NP-cPACNS also affects boys more commonly than girls [16]. Focal deficits include hemiparesis, hemifacial weakness, hemisensory loss, and fine motor skill loss. 40 % of the children with NP-cPACNS have headaches. Approximately 10 % of children present with additional diffuse focal deficits such as decreased cognition or behavior change [12]. Inflammatory markers including C-reactive protein (CRP) and erythrocyte sedimentation rate (ESR) are frequently normal. The endothelial cell marker von Willebrand factor (vWF) antigen appears to correlate well with disease activity [17]. The evaluation of potential prothrombotic abnormalities is mandatory; however, as mentioned before clots are more likely seen in adults than children as a cause for a stroke. Less than 50 % of the NP-cPACNS patients have an elevated protein level or evidence of leukocytosis on cerebrospinal fluid (CSF) analysis [12]. The role of the opening pressure remains uncertain.

CT scan has a low sensitivity for PACNS; however, it can help differentiate between NP-cPACNS and hemorrhage [7]. In cPACNS hemorrhage is not a common feature, unlike adult PACNS [18]. CT angiogram can be helpful although the specificity and sensitivity are less than magnetic resonance imaging [19, 20]. In NP-cPACNS the MRI reveals unilateral ischemic lesions in the large vessel distributions; most commonly the basal ganglia are affected (Fig. 40.2). Angiography best confirms vasculitis demonstrating unilateral stenoses and dilatations of the proximal segments of the anterior and/or middle cerebral arteries (Fig. 40.2). Beading and irregularity of stenoses can be present [19, 20]. In the recent years, vessel wall enhancement has shown to be helpful in PACNS [21, 22]. Gadolinium contrast studies of the affected vascular wall segment should, if possible, be requested, and in NP-cPACNS these reveal wall thickening and contrast enhancement due to wall inflammation.

Fig. 40.2

Neuroimaging of primary nonprogressive CNS vasculitis in a 6-year-old girl. Legend: axial magnetic resonance (MR) fluid-attenuated inversion recovery (FLAIR) images demonstrated unilateral lesions in the left basal ganglia (a). The corresponding diffusion-weighted images (DWI) confirmed the restricted diffusion into the lesion (also supported by ADC, images not shown) (b). The time-of-flight MR angiography revealed a stenosis with wall irregularity of the left middle cerebral artery (MCA) and anterior cerebral artery (ACA) (c) with evidence of gadolinium contrast enhancement (images not shown). Correspondingly the intravenous angiography demonstrated a vascular stenosis of both MCA and ACA with evidence of wall irregularity (arrow) in the MCA M1 segment (d)

Magnetic resonance imaging (MRI) and angiography studies are mandatory and should include T1, T2, fluid-attenuated inversion recovery (FLAIR), diffusion-weighted imaging (DWI), and apparent diffusion coefficient (ADC) sequences for parenchymal lesions [7].

Antithrombotic therapy is commenced in all children with NP-cPACNS, but regimens vary between centers. In children with high degree of vascular stenosis, heparin is frequently started, followed by long-term low-dose acetylsalicylic acid for secondary stroke prevention [23]. The role of adjunctive therapy with corticosteroids for 3 months remains controversial [24]. Nonprogression is confirmed on the repeat imaging at 3 months establishing no evidence of involvement of new vascular beds and resolution of contrast enhancement in the vascular wall in steroid-treated patients. Recurrent ischemic events are seen in 30–60 % of children. The long-term outcome remains to be systematically studied; it appears to be closely related to the location and extent of the ischemic lesion, stroke recurrence, and possibly the use of corticosteroids. Although not systemically performed or studied, early comprehensive rehabilitation, as practiced in adult stroke care models, appears to have striking benefits [25].

Angiography-Positive Progressive cPACNS (P-cPACNS)

Children with P-cPACNS commonly present with both focal and diffuse neurological deficits [12, 26]. P-cPACNS, like NP-cPACNS, predominately affect boys. Diffuse deficits develop insidiously in children with P-cPACNS and thus the diagnosis is frequently delayed in these patients. Children with P-cPACNS are commonly diagnosed at the time they develop focal deficits including hemisensory loss or fine motor skill deficits [12]. In addition difficulty in concentration, cognitive dysfunction, mood, and personality changes are present in these patients. Headaches are present in 95 % of P-cPACNS patients. The clinical and imaging pattern frequently found in angiography-positive P-cPACNS overlap with secondary CNS vasculitis of childhood. Thus, underlying systemic conditions should be carefully looked for [8].

In children with P-cPACNS, inflammatory markers and CSF cell count may be mild-moderately raised, but a normal CSF cell count or ESR does not exclude an angiography-positive CNS vasculitis [12]. In P-cPACNS parenchymal lesions on MRI can be ischemic and/or inflammatory and are commonly present in more than one vascular territory [19, 20] (Fig. 40.3). Bilateral MRI lesions are present in 25 % of P-cPACNS patients and more frequently have an asymmetric appearance. Characteristically, the angiography demonstrates vasculitis of proximal and distal segments of the cerebral arteries, typically involving multiple vascular beds [19, 20] (Fig. 40.3). Due to the proximal and distal involvement, different degrees of gadolinium contrast enhancement can be found in affected vessel wall segments [21]. Although isolated posterior circulation vasculitis is less common, the anterior circulation is more commonly affected. Conventional angiography provides additional information about the length and degree of stenosis and also visualizes collateral blood flow into the affected brain tissue identifying additional brain at risk. Required MRI sequences are identical to those performed in suspected NP-cPACNS [20].

Fig. 40.3

Neuroimaging of primary progressive CNS vasculitis in a 16-year-old boy. Legend: axial magnetic resonance (MR) fluid-attenuated inversion recovery (FLAIR) images demonstrated multiple, bilateral lesion involving both gray and white matter (arrow) (a). The corresponding diffusion-weighted images (DWI) confirmed the restricted diffusion of many lesions (arrow) (supported by ADC, images not shown) (b). The time-of-flight MR angiography revealed multiple stenoses and irregularities of vascular segments of all calibers involving both anterior and posterior circulation (arrow) (c) with evidence of gadolinium contrast enhancement (images not shown)

Treatment of P-cPACNS requires immunosuppressive therapy in addition to antithrombotic therapy. At the time of diagnosis, intravenous methylprednisolone pulses (30 mg/kg/day) are given at many centers for a duration of 3–7 days [12]. Subsequently daily oral prednisone in a dose of 2 mg/kg/day (max 60–80 mg) is used. Barron et al. first reported on the efficacy of cyclophosphamide in cPACNS [27]. In 2001 Gallagher et al. described cyclophosphamide efficacy in five children with P-cPACNS [26]. Intravenous monthly cyclophosphamide pulses given for 6 months, followed by oral maintenance immunosuppression while tapering the child off corticosteroids, are often used. The long-term outcome of children with P-cPACNS has not been systematically studied. Residual focal neurological deficits are often seen in this subtype [12].

Angiography-Negative, Small Vessel cPACNS (SV-cPACNS)

Small vessel cPACNS is increasingly recognized around the world. These children may present with severe encephalopathy, extensive focal and/or diffuse deficits, and/or seizure status and require a rapid, invasive evaluation including an elective brain biopsy [13, 14]. The mode of onset varies significantly from child to child. Some patients develop significant cognitive deficits over weeks and months, complain of constant headaches, or present suddenly with focal seizures. In contrast, other children have a rapidly progressive disease with a meningitis-like illness and other systemic features such as fever and fatigue [13, 14]. Seizures are common at diagnosis of SV-cPACNS and all seizure types are seen. Status epilepticus or refractory status epilepticus in previously healthy children mandates an evaluation for an underlying inflammatory brain disease, in particular for SV-cPACNS. Equally challenging is the differential diagnosis and includes demyelinating diseases, neuronal antibody-mediated inflammatory brain disease, and other less common conditions. SV-cPACNS has a female predominance in contrast to angiography-positive disease [13].

Although inflammatory markers are frequently abnormal in children with SV-cPACNS, the degree of abnormality varies between patients. Hutchinson et al. reported that three out of four children with SV-cPACNS had at least one abnormal inflammatory marker in the blood at diagnosis [14]. More importantly >90 % had an abnormal CSF analysis including increased CSF protein and/or cell count. Most commonly mild to moderate CSF lymphocytosis is seen.

The most common imaging finding in SV-cPACNS is inflammatory lesions in the white or gray matter [13]. Unfortunately a CT scan is not able to depict these inflammatory lesions and is therefore not informative in SV-cPACNS. Any MRI pattern can be seen in SV-cPACNS due to the ubiquitous presence of small blood vessels in the brain and spinal cord [28]. Most commonly inflammatory lesions are found in the subcortical white matter and cortical gray matter (Fig. 40.4). MRI abnormalities are present in the vast majority of SV-cPACNS patients at diagnosis. MRI lesions are best seen on T2/FLAIR sequences. In contrast to NP-cPACNS, ischemic lesions are very uncommon. Lesional gadolinium contrast enhancement is present in less than 50 % of children with active disease at diagnosis [13]. Meningeal contrast enhancement is equally infrequently seen; however, it is one of the few specific MRI finding of SV-cPACNS after infectious meningitis is excluded. In other IBrainD including demyelinating diseases, meningeal enhancement is not seen [29]. Autopsies have established the generalized character of small vessel vasculitis in contrast to the focal nature of disease suggested by detectable MRI lesions [30]. Repeatedly normal MRI studies have been seen in children presenting with status epilepticus, in whom the diagnosis of SV-cPACNS was confirmed on brain biopsy.

Fig. 40.4

Parenchymal imaging of six children with brain biopsy confirmed primary CNS vasculitis. Legend: axial magnetic resonance (MR) fluid-attenuated inversion recovery (FLAIR) images of children at diagnosis of angiography-negative, brain biopsy-positive primary CNS; vasculitis revealed several patterns of lesion including symmetrical or asymmetrical involving both gray and/or white matter (Fig. 40.3a). The majority of lesions were contrast enhancing, and none were diffusion restricted (images not shown)

All patients with SV-cPACNS have normal MRA and conventional angiography studies by definition [11, 13]. Other neuroimaging techniques have so far not provided additional diagnostic utility in SV-cPACNS. The next step in the diagnostic evaluation is an elective brain biopsy, which should be completed within 10 days from starting immunosuppressive therapy. The brain biopsy should preferably target lesions identified on MRI. However, these may either not be accessible or in functionally important areas [9]. In these children, non-lesional biopsies should be performed targeting the nondominant frontal lobe.

The diagnostic yield of elective brain biopsies performed for suspected inflammatory brain disease and other treatable conditions other than tumors in children was found to be 69 % (1996–2003) [31]. The review of brain biopsies in children with SV-cPACNS reveals intramural, inflammatory infiltrates consisting predominantly of lymphocytes and can also be detected in the perivascular space [9] (Fig. 40.5). Childhood CNS vasculitis is not characterized by vessel wall destruction, fibrinoid necrosis, or evidence of necrosis or granulomas as seen in other types of vasculitis as it is a predominantly lymphocytic vasculitis. Granulomatous infiltrates, which are frequently described in adult PACNS, have so far not been reported in children with cPACNS, and if present a granulomatous IBrainD or infection causing granulomatous infiltration such as Mycobacterium tuberculosis should be considered [9, 32].

Fig. 40.5

Brain biopsy histology of primary CNS vasculitis in a 7-year-old girl. Legend: hematoxylin and eosin stain of a lesional brain biopsy demonstrated a lymphocytic intramural infiltrated (arrow) in a small cerebral vessel. The lack of fibrinoid necrosis and granuloma is characteristic for primary small vessel CNS vasculitis in children. The immunohistological phenotyping revealed a predominant CD8 cytotoxic T-cell phenotype (not shown), magnification ×400

Children with SV-cPACNS require a combination of immunosuppressive therapy in addition to the mandatory therapy for seizure control, abnormal movements, or psychiatric symptoms [14]. In order to control the devastating brain inflammation and the resulting clinical features and to prevent disease-related damage, treatment should be initiated rapidly. Hutchinson et al. reported an open-label study of children with SV-cPACNS receiving a 6-month induction protocol consisting of corticosteroids (initial methylprednisolone pulses 30 mg/kg/day, max 1000 mg for 3–5 days followed by oral prednisone 2 mg/kg, max 60 mg/day with defined monthly taper) plus monthly intravenously cyclophosphamide pulses (500–750 mg/m², plus MESNA and hyperhydration) [14]. After 6 months children were switched to maintenance treatment with initially azathioprine but more recently mycophenolate mofetil (MMF). After 24 months 70 % of the children had no evidence of any functional neurological deficit as measured by the pediatric stroke outcome measure (PSOM), and the treatment was found to be effective and safe [14]. Case series from other centers supported the efficacy of cyclophosphamide and MMF [33, 34]. Most series document good recovery of neurological deficits. Anticonvulsive medications are continued beyond 24 months at many centers.

Differential Diagnosis

The differential diagnosis of childhood primary CNS vasculitis is wide and rapidly expanding. It includes secondary CNS vasculitis, non-vasculitic inflammatory brain diseases, and noninflammatory vasculopathies. It is useful to separately discuss the differential diagnosis of large vessel, angiography-positive cPACNS and small vessel, angiography-negative cPACNS. Secondary CNS vasculitis can affect any vascular segment.

Secondary CNS Vasculitis

The most common secondary CNS vasculitis occurs in children with infections (see Table 40.1) [35]. Most postinfectious CNS vasculitis similar to postinfectious vasculitis elsewhere in the body is typically a self-limited inflammatory disease and may only require a short course of immunosuppressive therapy. In contrast, organisms such as Streptococcus and Mycobacterium tuberculosis directly infect the vessel triggering a localized inflammatory host response leading to wall edema [36, 37]. The extent of vessel wall integrity loss and damage is determined by the specific interaction between pathogen and host. Vessel wall inflammation triggers a similar cascade as seen in childhood primary angiitis of the central nervous system, which includes endothelial activation, vascular stenosis, and secondary clot formation. In addition inflammation of the perivascular brain parenchyma is seen. Mechanisms contributing to the development of vasculitis following infection include immune complex deposition, cross-reactivity, and cytokine secretion [38]. One in five children with Mycobacterium tuberculosis meningitis develops secondary central nervous system vasculitis [36, 37, 39, 40]. MRI findings are nonspecific and may include ischemic lesions, white matter T2/fluid-attenuated inversion recovery inflammatory lesions, leptomeningeal enhancement, and tuberculomas [36]. Narrowing of the basal arteries, particularly involving the perforating branches of the middle cerebral artery, can be seen on angiography. Tuberculosis-associated central nervous system vasculitis requires both antimicrobial and immunosuppressive therapy [39].

Table 40.1

Causes of secondary CNS vasculitis in children

Infections

Bacterial infections

Streptococcus pneumoniae, Mycoplasma pneumoniae, Mycobacterium tuberculosis, Borrelia burgdorferi, Treponema pallidum, others

Viral infections

Epstein-Barr virus, cytomegalovirus, varicella zoster virus, parvovirus B19, enterovirus, hepatitis C virus, West Nile virus, human immunodeficiency virus

Fungal infections

Only gold members can continue reading. Log In or Register to continue