Pediatric stroke. Revascularization and reconstructive surgery in children with cerebrovascular disease

3. Aetiopathogenesis and risks

A pediatric stroke is heterogeneous in aetiopathogenesis. If in adults strokes are associated most frequently with atherosclerosis of brachiocephalic arteries (BCA) [26; 27], the etiology of strokes in children is diverse and complex [40]. In literature there is a quite exhaustive list of diseases and syndromes, which are fraught with a risk of cerebral ischemia in childhood, adolescence and youth. According to data of the American Heart Association & American Stroke Association (2012), half of all children, who had suffered a stroke, had risks [219].

Complexity and diversity of etiology imply a wide circle of specialists, who must keep an alert eye on strokes in their routine practice.

Heart diseases (congenital and acquired) present one of the most significant risks equal to about 20—30% of the causes of ischemic strokes in childhood [161]. A combination of left-heart embolisms (or paradoxical embolism in right-to-left cardiac shunt) and cardiac decompensation is important in pathogenesis of cardioembolic variant of an ischemic stroke [1; 7; 22; 52]. There is a description of cases of paradoxical embolism into the cerebral vessels of children and young people in the background of an atrial septal defect, open foramen ovale, in cases of arteriovenous malformations of pulmonary vessels and neurocutaneous syndromes [280]. While this problem was in focus, the attention was again attracted to minor cardiac abnormalities. Regarding patients with vague etiology of stroke, it is, primarily, recommended to rule out sources of hidden or paradoxical embolism as an open foramen ovale, a mitral valve prolapse and an atrial septal aneurysm [1; 52; 56; 38]. The literature data state that imaging reveals «silent» brain infarctions in 25% of patients with mitral stenosis. Also, clinically «silent» ischemic lesions of brain tissue are found in 20% of newborns with heart diseases on a pre-surgery stage and in 17.4% – on a post-surgery stage [63; 181; 221]. Presently, several studies were held with the attempted prognostication of strokes in infants with congenital heart diseases. A significant role of duration and an intensity of hypoxia in newborns, resuscitation procedures, prematurity and duration of waiting for surgical intervention were indicated as ACVD risks at pre-surgery and post-surgery stages [62; 112; 161; 181; 183; 210].

Cardiac arrhythmias are considered to be a very rare cause of strokes in childhood and youth, as opposed to adults. Nevertheless, it should be kept in mind regarding children with hyperthyreosis, rheumatic heart diseases, after surgical interventions and in the structure of Kearns-Sayre syndrome.

Cardiac myopathy as a manifestation of systemic diseases occurs in congenital myopathies (Duchenne, Becker, Emery-Dreifuss, etc.), Friedrich’s ataxia, mitochondrial diseases. With this pathology, both embologenic and hemodynamic variants of an ischemic stroke are possible. In some cases, a myocardial infarction and a stroke may develop simultaneously, which points at the similarity of pathogenetic processes leading to inadequate perfusion [7].

Hypercoagulation states are presently considered to be the most common causes of ischemic strokes in childhood – their contribution reaches 87% [3; 16; 138; 144; 170; 265]. However, a universally acknowledged screening protocol for thrombophilic state in a child with CVD has not been developed yet, and some researchers dispute the role of multigenic thrombophilias as risks of strokes and TIAs in children [82; 121; 142; 184; 194; 282].

During the last decade a large number of thrombophilic mononucleotide genic polymorphisms were described. Carriership of proaccelerin, prothrombin, plasminogen activator inhibitor and fibrinogen is considered to be most significant clinically [61; 138]. F5 genotypes: 1691 G> A and AA (Leyden mutation) as well as F2: 20210 G> A and AA are currently the only ones, whose prothrombotic effect is acknowledged in newborns. Besides, there is a good reason to suppose that with the carriership of thrombogenic mutations and polymorphisms in children the risk degree differs according to their age [5; 56; 138].

Presently, there are no doubts regarding the connection of hyperhomocysteinemia and MTHFR677C> T mutation with cerebrovascular and cardiovascular diseases [5; 16; 61; 280]. Homocystein acts as a prothrombotic factor due to activation of coagulation factors XII and V, increase in tissue factor expression and suppression of thrombomodulin expression. Besides, the rise of homocystein in blood leads to vascular endothelium damages, which reveals in neurotoxic and proatherosclerotic effects and contributes to the emergence of resistance to activated protein C [3; 31; 66; 267; 276; 290]. During the latest years, the role of hyperhomocysteinemia in damaging the vascular walls was proven as well as its prothrombotic and pro-atherosclerotic effects and its effect on the vasomotoric regulation [3; 32; 290].

The MTHFR enzyme gene provides for homocystein metabolism with the participation of a folic acid. The greatest practical significance belongs to a mononucleotide replacement of cytosine with thymine at position 677 of gene, thus leading to a replacement of alanine amino acid residue with valine in the catalytic core of methylene tetra hydro folate reductase enzyme (MTHFR). Individuals, homozygous by this allelic mutation, display a decrease of the enzyme activity by 60—70%, and heterozygous – by 35% [28; 159; 275]. It should be noted that all available data concern either fundamental aspects of the pathology study, or the population of adult patients.

According to literature data, children are noted to have a positive correlation relationship between the incidence rate of strokes (especially, in boys) and C677T polymorphism. The combination of several variants of genes is accompanied by a progressive rise of homocystein level in blood, and it increases the risk of a CVD [61; 142; 229; 255; 269; 289]. At the same time, a series of genic polymorphisms controlling the folate cycle activity showed their protective function regarding the cerebrovascular pathology in Asian and European populations [82; 142]. Despite the indirect relationship between pheno- and genotype of hyperhomocysteinemia, the researchers’ opinion is unanimous: determination of the homocystein level and the state of folate cycle genes must become an insatiable component of diagnostic suite in this group of patients, especially, in boys (class II, recommendation level B) [280].

In the literature there is a description of singular clinical cases of BCA thrombosis in children. There are no general statistics on CCA, ICA and MCA occlusions in pediatric population neither in national, nor in foreign literature. In 1951 M. Fisher was the first to describe the post-thrombotic occlusion formation steps in ICA of adults with hemodynamic and embolic mechanisms of an ischemic stroke, having analyzed the angiograms of patients with CVDs [96]. Post-thrombotic occlusions of major arteries in the heads and necks of children account for from 13 to 37% in the structure of CVD causes.

In childhood there are more than enough initiating agents capable of worsening the hemorheologic situation or decreasing the athrombogenic properties of a vascular wall. The adverse course of the post-natal adaptation period, infection, microtraumatization and metabolic disorders can act as triggering factors and lead to an acute cerebral ischemia in the background of the carriership of mononucleotide polymorphisms in thrombophilic spectrum genes and in the genes, which control the activity of folate cycle enzymes. In children with CVDs (especially, when the onset is in the perinatal period), it is recommended to search for prothrombotic mutations even, when other causes of ACVD are identified (class IIa, recommendation level C) [107; 280]. At the same time, the detection of thrombophilic polymorphisms is not an absolute factor inevitably leading to thromboses. Attention should be paid to the quantity of revealed mononucleotide mutations, the fact of homozygous carriership, the variants of genes – genic combinations and their phenotypic manifestations [5; 121; 286].

If the researchers have no doubts about hyper-homocysteinemia presenting a risk of thrombi formation at a non-typical age, the atherothrombotic variant of CVD is a casuistry for pediatric practice. Such variants are described in singular cases and associated with proven, genetically determined, dyslipidemic syndromes [165; 167; 189], most of which proceed asymptomatically [54]. Nevertheless, the selective screening and the lipid metabolism monitoring are recommended for patients with a family history of an early onset of vascular diseases and for the ones with an unclear cause of the stroke [111; 239].

The significance of the infection process as the releasing factor with the developing acute cerebrovascular insufficiency in the background is great both in newborns (up to 17.6% among all causes) and in elder patients (up to 40.7%) [60; 63; 223; 226; 280]. The clinical study showed the significance of a short duration (up to 4 weeks) and the fact of minor infections as an independent risk, which both enhance the probability of an ischemic ACVD 4.6 times as much [43; 240]. When analyzing the strokes of an unknown etiology, it was noted that shortly before the stroke children had varicella 3 times more often than in the population; the probability of a stroke also remained high within the first four months after varicella [157; 160]. Besides, the study held in 2006 showed that, in children with the earlier revealed immunodeficiency, the risk of the recurring stroke grew 20.9 times as much and correlated directly with the level of white blood cells during the acute period of the disease. A chronic infection and the immunodeficiency are supposed to make their own contribution to the development of recurrent incidents of CVD in children in the same way, as in adults [108].

It cannot be ruled out that an infectious process flows in the nervous system by the mechanism of vasculitis. Presently, the VIPS study (The vascular effects of infection in Pediatric Stroke Study) is held with the hypothesis stating that the presence of an infectious agent triggers an endothelial dysfunction, a systemic inflammatory process, which, combined with insufficiency of connective tissue and prothrombotic readiness, leads to damages of the vascular wall, its dissection, thrombogenesis, luminal occlusion and cerebral ischemia [13; 101; 127]. Later on, an embolism may occur from the artery dissection site as well as the hemorrhagic transformation of an ischemic lesion [13; 16; 101]. Also, apart from vasculitis, some hemorrhagic complications caused by coagulopathy may occur in severe somatic diseases [54; 55]. Presently, there are no such distinct diagnostic criteria of cerebral vasculitis in children, which could provide grounds to assert with confidence that it was this disease that had caused the CVD [13]. It is proposed to keep in mind that vasculopathy may be the most probable etiology of a CVD in all cases of TIA, and that it happens always in pediatric or young patients, especially, in the absence of evident risks [22; 55; 168].

During their early period of life, in children with CVD one cannot ignore the pathological course of the prenatal period, which could act as one of the initiating agents [105; 280]. Hemoreologic situation deterioration, arterial blood pressure measurement, endothelial dysfunction, systemic inflammatory response and other pathological processes triggered by a perinatally caused hypoxia initiate a hypercoagulation state and a thrombosis of various localization. In its own turn, the cascade microthrombogenesis mechanism affects the perfusion situation at the thrombosis site, maintaining the hypoxia, initiating the necrotic and apoptotic mechanisms of nerve cell death. Besides, in infants with combined impairment of central nervous system and extra-cerebral pathology the adverse course of the period of adaptation to new living conditions may become the basic cause of inadequate blood supply to cerebral structures with the decrease of the cerebrovascular reserves [13; 19; 90; 223; 230]. Certainly, all by itself, the process of adaptation of a fetus and a newborn to new conditions of existence is physiological, but the mother’s diseases occurring before and during the pregnancy increase the probability of developing both ischemia and hemorrhages into cerebral structures of newborns and infants at all pre- and post-natal adaptation stages [25; 90; 140; 163]. Additional risks include intra-uterine infections, head and neck injuries, systemic bowel diseases, autoimmune diseases, water depletion, infertility in the mother’s history, premature rupture of fetal membranes, the mother’s pre-eclampsia [22; 34; 219]. Burdened perinatal anamnesis occurs in every fourth patient and, it is associated more with neonatal and fetal stroke than with a CVD in an elder age group [34; 223].

Radiation-induced vasculopathies with subsequent vascular stenosis or occlusion more often occur in patients with brain tumors. A recent study showed that after radiation therapy courses the radiographic signs of strokes were noted approximately in 6% of children with tumors of the central nervous system [92].

One of the causes of brain blood supply disturbance in children is a cerebrovascular pathology. Presently, many authors come to the conclusion of an imperative need in a more detailed examination of children for presence of cerebrovascular diseases. In the situations, which require a compensation by means of the blood flow redistribution (e.g., during a prolonged head tilt), the decisive role may be played by specific anatomic features of major arteries of the head and the neck as well as of the basilar vessels, which can often be found even in healthy people (non-closed Willis’ circle, hypoplasias of individual artery segments or of whole arteries, stenosis / occlusion of arteries, pathological deformation of ICA) [7; 12; 30; 108].

The prevalence of ischemic disturbances, which, as a rule, are preceded by a chronic cerebrovascular insufficiency, in the structure of the young age stroke requires a comprehensive study of age peculiarities in the cerebrovascular system functioning. In principle, the brain blood supply physiology in children does not differ from that of adults. For instance, a brain weight is only 2% of the total weight of man, but it receives about 20% of the blood volume during the cardiac output and consumes 20% of oxygen. An adequate cerebral blood flow sufficient for sustaining a normal life activity of the brain is, on the average, 45—60 mL/100 g of brain tissue per minute. The proper brain activity depends on regular and adequate blood supply due to its high metabolic activity and lack of any significant energy reserves. The state of the cerebral blood flow insufficiency known as a cerebral ischemia may be either acute, or chronic and may have either local or widespread nature. When the blood flow decreases to 20—35 mL/100 g/min., the electrical activity of the brain drops, but the brain tissue changes are reversible. This state is often defined as a «penumbra» – an ischemic penumbra. When the blood flow decreases to 10—15 mL/100 g/min., the changes developing in the brain tissues become irreversible. This process leads to the death of cells (brain infarction). Timely restored adequate cerebral blood flow is enormously important in treatment of a specific group of patients with a chronic cerebral ischemia caused by a pathology of major cerebral vessels [92]. The study of cerebrovascular disturbances may turn to be a connecting link between a cerebrovascular pathology in children and the subsequent development of strokes in adults.

The anatomy of brachiocephalic arteries varies in individuals. Both in children and adults, the Willis’ circle is formed correctly only in 18—20%. Many authors think that the anatomy of cerebral vessels does not depend on gender, age or race [89; 150]. In the opinions of other authors, in children the specific anatomic and physiological features of BCA and intracranial arteries take shape, mainly, in the younger age group. For instance, the collateral vessels between ECA and ICA via a. ophtalmica are formed only in 17—22% of children [173]. The morphometric parameters of ICA grow in school-aged children from 7 to 18 years old. The morphometric parameters of common carotid arteries grow leapwise. In girls, the maximum growth occurs earlier than in boys [6].

More than a century ago in literature there were already mentions made of specific anatomical features of ICA structure – tortuosity, kink and looping. Many anatomists described bent, deformed ICA. W. Coulson (1852) was, probably, the first, who mentioned the ICA looping describing it as «a pulsating tumor on the neck». The connection between the pathologic deformations and the high risk of the CVD development was first noted by M. Riser et al. in 1951 [218]. F. McDowell et al. (1959) reported 20 cases of pathological tortuosity of ICA and related blood flow disturbances, when studying 68 patients with cerebrovascular pathology among adults over 30 years old without any proofs of occlusion found, which was supported by angiography [92]. H. Metz et al. (1961) held a retrospective survey of 1000 angiograms and found 161 cases of kinks and looping on ICA, including children below 10 years old [182]. In 1965 after having described 2,453 angiograms J. Weibel and W. Fields revealed a high occurrence of this specific feature of anatomical structure in the population – up to 10—15% cases [274]. N. Sarkari et al. (1970) first described 8 children with the pathological tortuosity of ICA combined with a chronic cerebral ischemia, of which 7 children were below 10 years old [228].

The development of contemporary diagnostic methods led to more frequent detection of ICA deformations in a population. According to US test data, the detection rate of ICA tortuosity reaches 25—30% in adults and 43% in children (R. Hobson, 2004). According to data of other authors (E. Ballota et al. (2005), G. La Barbera et al. (2006), W. Perdue et al. (1975), C. Togay-Isikay et al. (2005)), pathological deformations occur in 10—40% of cases depending on the population surveyed.

In the opinion of W. Fisher (1982), the surgical treatment of cerebrovascular insufficiency in children is required in cases, when the ICA changes may lead to a progressing neurologic deficiency. The author also states that cerebrovascular insufficiency in children presents a serious problem, as an inadequate perfusion of brain tissue leads to irreversible changes with the subsequent formation of a neurologic deficiency. In his opinion, the most frequent causes leading to cerebrovascular insufficiency are the following ICA changes: 1) congenital pathological tortuosity (kinking), 2) post-traumatic changes with thrombosis, aneurysm and embolism, and 3) thromboses [97].

In children, CVDs may be also caused by: the moya-moya disease, a fibro-muscular dysplasia, an artery dissection [8; 20]. All the above-mentioned diseases lead either to stenosis or luminal occlusion [213].

Fibro-muscular dysplasia is one of the causes of a pathological deformation of the ICA and a pediatric stroke. The pathomorphology of the fibro-muscular dysplasia is characterized by the emergence of hyperplasia sites on the connective tissue, an irregular atrophy of muscular fibers and degenerative changes of vascular walls. All this leads to formation of stenosis sites interspersed with post-stenotic aneurysms. This variant of impairment of brachiocephalic vessels may be isolated and generalized, more frequently unilateral, although a bilateral impairment is not ruled out [211].

Fig. 2: Types of pathologic deformations according to SCT AG data: A – Type 1 septal stenosis (angle over 60°) (arrow indicated); B – Type 2 septal stenosis (angle from 30 to 60°) (arrow indicated); C – Type 3 septal stenosis (angle less than 30°) (arrow indicated).

The permanent effect of the force vector of arterial blood pressure on dysplastic vascular wall was discussed earlier as the most intensive initiating agent forming a pathologic tortuosity of the ICA [145]. In 1961 H. Metz offered his own classification of ICA deformations. The author supposed that the intensity of the process depended on dysplastic changes of the vascular wall and vessel configuration and divided septal stenoses of the ICA into three types:

Type 1: artery kink at an angle over 60°(Fig. 2A);

Type 2: artery kink at an angle from 30 to 60°(Fig. 2B);

Type 3: artery kink at an angle less than 30°(Fig. 2C) [182].

In 1974 O.V. Voronin singled out three groups of pathologic deformations: looping; acute angle formed by arteries; artery tortuosity without distinct angulation. In 2001, based on 250 case follow-ups, P.O. Kaznanchyan et al. divided pathologic deformations into the following basic groups:

1. C- or S-shaped tortuosity;

2. artery elongation and artery kink at an angle of less than 90°(angulation) causing a local stenosis of a major artery;