PREDICTIVE METHODS FOR THE ASSESSMENT OF THE RISK OF THE APPEARANCE OF PERINATAL CEREBRAL LESIONS IN THE FORMATION OF PROGNOSTICS OF RESIDUAL DISORDERS IN THE CENTRAL NERVOUS SYSTEM

Complications affecting central nervous system (CNS) such as hypoxic ischemic lesion, often are responsible for the delaying of neurological development. Many authors consider that it occurs mostly before and during delivery [1, 2]. Fewer than 36% of deaths in children under five years of life occur in the neonatal period and are caused by premature birth, birth asphyxia and infections. Some authors note that cerebrovascular accidents (CVA) are one of the most common complications of hypoxia and ischemia in the first week of life [1]. The authors of a Canadian Guide on conditions related to CVA considered that although neurological development is often abnormal the cause remains unclear and must be addressed to the perinatal period [2, 3].

Perinatal cerebral lesions (PCL) can have an impact on child’s neuro-psychical development which supposes medico-social and economic burdens. The severity of the lesion at birth process, as well as early and accurate applied therapeutic measures are decisive for prognosis. Consequences of PCL determined from 20 – 40% up to 60% of affected children who survive an injury of the brain. Several studies showed that the consequences of the cerebral hypoxia vary from mild behavioral deficits to severe pathology manifested with mental retardation and/or cerebral palsy (CP) and epilepsy (EP) [4]. In recent years, numerous studies aim to find measures of prevention, early diagnosisand treatment of cerebral lesions by neuroprotective agents [5-8]. Efforts made to understand the pathogenesis underlying residual pathology (RP) of the CNS [9-13]. Some authors argue that, despite an adequate management of hypoxic-ischemic neonatal encephalopathy (NE), in much cases perinatal asphyxiation cannot be prevented [14].

The consequences of the PCL in most cases are aggravated by delayed diagnosis and ineffectiveness of early administered treatment. Current researches in neuroscience are directed toward identifying models of preventing the damage and death of nerve cells due to various pathological factors. Despite researches conducted over the past two decades, there are no clear principles of PCL prevention.

Scope of the study: the research of risk factors, including highlighting of etiological variables, early clinical and paraclinical manifestations, which allow to develop the models of prognosis of residual pathology of CNS in children with perinatal cerebral lesions.

Methodology of research: there observed 1036 view (L1) full-term children, who suffered hypoxicischemic neonatal encephalopathy (NE), distributed according to the degree of severity of perinatal cerebral lesion (PCL): 422 cases (L1A) – I grade, 310 cases (L1B) II grade, 304 cases (L1C) III grade. The control group (L0) comprises 344 “conventional healthy” children.

All the children were supervised from 1 – 3 up to 60 (48 ± 12) months (Table 1).

All cases were assessed through historical examination, clinical, imaging and electroencephalographic investigations. To elucidate pathogenetic and prognostic peculiarities of the PCL, in a group of 100 children selected from the main group (for 25 of each I, II, and III grade and 25 of “conventional healthy” children), at the ages of 1 – 3 months, to determine in serum the Brain-derived neurotrophic factor (BDNF) and Ciliary neurotrophic factor (CNTF) by Enzyme- Linked Immunosorbent Assay (ELISA). Was applied statistic software set SPSS v1.8, QUANTO v1.2, Review Manager (RevMan) vl. 5.1, GMDR software Beta 0.9, using the following test methods: arithmetic mean, standard error, relative frequency, correlation coefficient, Student test, χ2 (chi square) test, ANOVA test, Mann-Whitney test, logistic regression.

Table 1. Distribution of children in the study groups according to the degree of severity of PCL, based on retrospective data (abs.,%).

Table 2. Description of logistic regression

The results of the logistic regression model of the relationship between the lots of independent variables and a dichotomous dependent variable (Table 2).

The probability of the event for each item of study can be calculated by the following formula: , where e – is a mathematic constant = 2,72.

Example: case in which all variables are positive we get the following equation of prognosis:

Y = -2,346 + (1,633 × 1) + (0,997 × 1) + (1,377 × 1) + (2,15 × 1) + (2,121 × 1) = 5,932

Probability of development of RP in this case is 0,997 or 99,7%.

Explanation: the child who are present all the risk factors has a likelihood of developing the disease of 0,997 (99,7%).

Another example, in which the child have only first 4 prognostic criteria without the asphyxia we get the following equation:

Y = -2,346 + (1,633 × 1) + (0,997 × 1) + (1,377 × 1) + (2,15 × 1) + (2,121 × 0) = 3,811

Probability of development of RP in this case is 0,978 or 97,8%.

The case in where present intrauterine hypoxia and pathologic course of pregnancy we get the following equation:

Y = -2,346 + (1,633 × 1) + (0,997 × 0) + (1,377 × 1) + (2,15 × 0) + (2,121 × 0) = 0,664

Probability of development of RP in this case is 0,660 or 66,0%.

Thus, depending on the number of variables, we can calculate the risk of development of RP.

Obtained results. Health status of parents before conception, especially mother’s health, is important for health of future children. In present study by selfassessment were determined as healthy 1249 women (91,2%; 95CI 90,43-91,97) of 1370 followed children, maximal share was in mothers of children from L0 – 317 cases (94,9%; 95CI 93,7 – 96,1), and minimal in mothers of children from L1C – 258 (84,9%; 95CI 82,84 – 86,96) (χ2=22,0, gl=3, p<0,001) (Table 3).

Most health problems during pregnancy were noted in mothers from L1C – 228 (75,0%; 95CI 72,52 – 77,48), followed by mothers from L1B – 197 (63,5%; 95CI 60,77 – 66,23), and mothers from L1A – 246 (58,3%; 95CI 55,9 – 60,7). Mothers from L0 were “conventional healthy”, 51 (15,3%; 95CI 13,3 – 17,3) suffered from anemie I grade, and 25 (7,5%; 95CI 6,1- 8,9) suffered from pyelonephritis.

Evolution of pregnancy was normal in 235 cases (17,2%; 95CI 16,2 – 18,2), majority of mothers of children from L0 – 214 (15,6%; 95CI 14,62 – 16,58). Most prevalent complication of pregnancy was early toxicosis of pregnancy – in 740 cases (54%; 95CI 52,65 – 55,35), followed by threat of miscarriage – in 459 cases (33,5%; 95CI 32,22 – 34,78) and late toxicosis of pregnancy – in 333 cases (24,3%; 95CI 23,14 – 25,46). The prevalence of pathologic evolution of pregnancy in studied group was higher in L1B and L1C groups (Figure 1, Table 3).

Trauma during pregnancy have suffered 22 women (1,6%; 95CI 1,3 – 1,9). The smallest share of trauma was registered among the mothers from L0 – 2 (0,6%; 95CI 0,2-1,0), and highest share among the mothers from L1C – 12 cases (3,9%; 95CI 2,8-5,0) (χ2=14,6, gl=3, p<0,01).

Pathological movements of fetus were registered in 157 cases (11,5%; 95CI 10,6 – 12,4). The better results were registered in the group of healthy children (χ2=122, gl=3, p<0,001).

One of more frequent registered problem of pregnancy was intrauterine hypoxia – 982 cases (71,7%; 95CI 70,48 – 72,92), with evident and statistic reliable difference between the children with outcome to PCL – 974 cases (71,1%; 95CI 69,88 – 72,32) and healthy – 8 cases (2,4%; 95CI 1,56 – 3,24) (p<0,000).

Another most prevalent problems before birth were: pathological evolution of pregnancy (early of late toxicosis, threat of spontaneous abortion, toxemia of pregnant women etc., p<0,004), pathologies of placenta (placenta praevia, placenta abruptio, polyhydramnios, p<0,000) and amniotic membranes (chorioamnionitis, early or late membrane rupture, p<0,000), different pathologies of umbilical cord (circulating or strangulation, short umbilical cord etc., p<0,006).

Was also important postnatal diseases, including postnatal asphyxia of different grade, which were in 535 newborns (39,1%; 95CI 37,78 – 40,42). The share of this aggravating factor was higher among children who subsequently developed neurological disorders – in 533 cases (51,4%; 95CI 49,85 – 52,95) (p<0,003, rxy=0,714). Associated perinatal trauma was registered in 235 newborns (17,2%; 95CI 16,18 – 18,22), most prevalent in children from L1C – 172 cases (56,6%; 95CI 53,76 – 59,44), (p<0,001) (Table 3).

Status of the child immediately after birth is assessed by the Apgar score. At one minute after birth, Apgar score was 5,4±0,1 points, with minimal value 1 and maximal value 9 points. The lowest average Apgar score was recorded in L1C – 2,4 points (95CI 2,36 – 2,44), while the highest one in L0 – 7,7 points (95CI 7,66-8,36). Over 5 minutes after birth, the average value of the Apgar score has increased compared to the immediate period after birth, being average 6,9 points (95CI 6,85 – 6,95), with values ranged from 2 to 10 points. As in the case of data recorded at 1 minute, the lowest average score over 5 minutes was recorded in L1C – 4,5 points (95CI 4,45 – 4,55), while the highest one in L0 – 8,6 points (95CI 8,56 – 8,64).

Have been registered anthropometric parameters at birth, i. e., weight and head perimeter. The average weight of children included in the study was 3381,5 g (95CI 3368,8 – 3394,2), in children from L1C was registered lowest average value was 3300,4 g (95CI 3264,7 – 3336,1). The average head perimeter of children included in the study was 33,7 cm (95CI 33,68 – 33,72), with registered lowest average value in children from L1C – 33,4 (95CI 33,32 – 33,48) cm.

Significant statistic difference between group was noted – p<0,001. Children born with gestational age 37 – 38 weeks were more prevalent in L1C, but born at more than normal term (gestational age more than 40 weeks) – in L1A and L1B.

The presence of syndromes suggestive for a neurological pathology was important in determining the severity of PCL and distribution of children in groups for surveillance. The most common symptoms in this regard were mentioned in children from L1C (Figure 2, Table 3). The duration of clinical manifestations is one of the most important criteria in assessing the severity of the disease. Thus, in the children from L1A clinical symptoms were extended during the 24/36 hours, in L1B 72/94 hours, and in those in theL1C more than 1 week (Table 3).

Fig. 1. Frequency of problems that determined the pathological evolution of pregnancy,at the mothers of the children enrolled in the study (n = 1370,%)

Note: EG – early gestosis, IA – imminence of abortion, LG – late gestosis, NF – nephropathy, HA – hypertension,

FVI – frequent viral infections, OW – overweight, SR – seizures, HC – hypercholesterolaemia, PhE – physiological evolution.

Table 3. Noxious factors with Significance for Development of Residual Parodies of CNS in Children.

Fig. 2. Clinical manifestations encountered in the newborn period suggestive of RP development, (%)

Neurosonografic examination allowed the detection of anomalies in brain tissue, manifested with: stasis or peri-ventricular edema, peri-ventricular pseudo cysts, peri-ventricular dilation of ventricles of different severity, high neurosonographic density etc. Most of the changes have been determined in the children from L1C, from the difficulties of imaging of cerebral tissue – 114 cases (37,5%; 95CI 34,72 – 40,28; p<0,001) and dilation of different severity of cerebral ventricles – 273 cases (89,8%; 95CI 88,06 – 91,54;

p<0,001), up to stasis and peri-ventricular edema – 214 cases (70,4%; 95CI 67,78 – 73,02; p<0,001), cerebral hemorrhage – 6 cases (2%; 95CI 1,2 – 2,8; p<0,05), high neurosonographic density of choroid plexus – in 138 cases (45,4%; 95CI 42,54 – 48,4; p<0,001), etc. (Figure 3). Neurosonographic changes correlates with severity of PCL: L1A (rxy=-0,24), L1B (rxy=-0,53) and L1C (rxy=-0,78).

Electroencephalographic features (EEG) in newborn children from L0 were normal bioelectric activity.

In 139 cases (32,9%; 95CI 30,61 – 35,19; t=6,3547) from L1A in small age bioelectric changes manifested as excitatory activity (excess of sharp waves and peaks of high frequency) and signs of bioelectric immaturity of the brain (slow waves, especially deformed theta waves combined with delta waves). Thus non-specific features correlates with minimal residual disorders (rxy=-0,22; p<0,001). In small age in majority of cases in L1B were registered the following features: excitatory activity in 50 cases (16,1%; 95CI 14,01 – 18,19; t=6,0108; p<0,001); slow waves in 34 cases (11%; 95CI 9,23 – 12,77; t=0,4916; p>0,05); sharp slow waves in 18 cases (5,8%; 95CI 4,47 – 7,13; t=7,3829; p<0,001) and slow isolated activity in 28 cases (9%; 95CI 7,37 – 10,63; t=6,4798; p<0,001), that suggests the presence of low convulsive threshold, with risk of PE development in future (rxy=-0,42). In L1C were registered the following EEG abnormalities: recordings with reduced amplitude in 147 cases (48,4%; 95CI 45,53 – 51,27; p<0,001), from which slow waves in 37 cases (12,2%; 95CI 10,32 – 14,08; p<0,01); sharp slow waves in 56 cases (18,4%; 95CI 16,18 – 20,62; p<0,01); isolated slow activity in 16 cases (5,3%; 95CI 4,02 – 6,58); diffuse slow activity in 18 cases (5,9%; 95CI 4,55 – 7,25; p>0,05), and also suppression burst activity in 25 cases (8,2%; 95CI 23,42 – 9,78; p<0,001) and focal spike-and-wave activity in 16 cases (5,3%; 95CI 4,02 – 6,58; p<0,05), which confers significant risk of PE in future (rxy=-0,72).

Fig. 3. Some ultrasound issues encountered in children with PCL in the first month of life.

Note: NA – normal appearance, CSNCV – cerebral substance not clearly visualized, DLV – dilated lateral ventricles (lg – light, mod – moderate, sev – severe), WISS – Wider inter-spherical space, ISS – enlarged subarachnoid space, BVC – blurred ventricular cavities, PVS – periventricular stasis, PVE – periventricular ecobodies, PVCs – periventricular cysts.

In the study, clinical and paraclinical manifestations have established a relationship between clinical diagnosis and PCL severity; much worse are clinical manifestations, the more severe is the diagnosis.

Mann-Whitney criterion confirmed this relationship, the significance of the case being – mY=0,0021.

Simultaneously with neuro-psychological testing in children with PCL as well as in healthy children was determined the serum levels of neurotrophic factors BDNF and CNTF. Obtained results are characterized by statistic significant variation of concentrations of neurotrophic factors. Highest values of BDNF and CNTF were determined in L0, and lowest levels in children with severe cerebral lesions (L1C). Was established strong indirect statistic correlation between serum level of BDNF (rxy=-0,918, p<0,001) and CNTF (rxy=-0,921, p<0,001) with grade of severity of PCL. At the same time, it has a strong direct correlation between BDNF and CNTF at age 3 months (rxy=0,939), which confirms implication of both neurotrophic factors in process of neurologic development and degeneration. Application of these tests in children with PCL facilitates probability revealing of RP, increasing several times the probability for each condition, which means improving the diagnosis in patients suspected for RP of the CNS (Figure 4).

We also have strong correlations observed between low values of neurotrophic factors (BDNF and CNTF) and structural lesions of the brain (rxy=-0,72, p<0,001), confirmed by severe diagnosis as CP and EP (rxy=-0,59, p<0,01). Similarly, indirect correlations were obtained between serum concentrations of neurotrophic factor sat the age of 1 – 3 months and EEG changes, L1B: BDNF (rxy=-0,24) and CNTF (rxy=- 0,22); L1C: BDNF (rxy=-0,72) and CNTF (rxy=-0,74), suggesting evolution into severe pharmacologically resistant forms of EP.

Comprehensive investigations by predicting more parameters have allowed a mathematical analysis of evaluation of relative risk for the development of the RP. Were obtained the following results: in children from L1A: RP+=1,2–1,6; L1B: RP+=1,7–2,5; L1C: RP+=2,5–3,5. These data provides the opportunity for the appreciation of the risk of development of RP in children with PCL, thus facilitating the selection of the method of assessment of neurological development in long-term perspective, and allow adequate psychological approach, justified in order to improve the prognosis and social classification of the subjects in this category.

DISCUSSIONS

Hypoxic ischemic neonatal encephalopathia (HINE) caused by decreasing of O2 concentration in blood and disorder of cerebral blood flow, capable provoke encephalopathic process expressed clinically by neurologic syndrome [13, 15], with long-term consequences in the process of neurological development [16]. Documenting the neurological symptoms depending on the accumulated scores at birth, timely evaluation of the prognosis of infants with HINE allowed an explanation of the results obtained in long-lasting prospective study [17]. The most important role in long-term prognosis of RP of SNC have severity and duration of hypoxia. The authors claims that infants with hypoxia often are born with metabolic disorders or mixed acidosis with a low pH (< 7) in the blood obtained from umbilical cord, Apgar score being small (< 3 points) from 0 to 5 minutes after birth. Sometimes, in severe cases, from 12 to 36 hours, HI episode evolved into NE [16]. Often it is found the association between pH of blood from the umbilical cord at birth and long-term neurological consequences [18].