Acta Paadiatr 81: 002-7. 1992

Hyperbilirubinemia, hypocarbia and periventricular leukomalacia in preterm infants: relationship to cerebral palsy RS Ikonen, MO Janas, MJ Koivikko, P Laippala and EJ Kuusinen Department of Pediatrics, Tampere University Hospital, Tampere. Finland

Ikonen RS, Janas MO, Koivikko MJ, Laippala P, Kuusinen EJ. Hyperbilirubinemia, hypocarbia and periventricular leukomalacia in preterm infants: relationship to cerebral palsy. Acta Prediatr 1992; 81:802-7. Stockholm. ISSN 0803-5253 This study comprised 103 preterm infants with a gestational age less than 33 weeks who were born in Tampere University Hospital and-who were followed up to two years of age. Sixty-four perinatal variables were compared to ultrasound findings in the neonatal period and neurologic handicap at the age of two years. Duration of hypocarbia (PC02< 30 mmHg) during the first 72 h and hyperbilirubinemia (the mean level of serum total bilirubin) at three days of age were independently and significantly related to periventricular leukomalacia, but not directly to cerebral palsy. The only perinatal variables related independently and significantly to cerebral palsy at two years of age were periventricular leukomalacia and ventriculomegaly. According to these results, periventricular leukomalacia was the main predictor of cerebral palsy in preterm infants. In addition to hypocarbia, hyperbilirubinemia may also be involved in the pathogenesis of extensive (severe cystic) periventricular leukomalacia. 0 Cerebral palsy, hyperbilirubinemia, hypocarbia, ischemia, periventricular leukomalacia, preterm infant RS Ikonen, Department of Pediatrics, Tampere University Hospital, SF-33520 Tampere. Finland

The toxicity of high bilirubin levels to the newborn brain is generally accepted. The precise biochemical mechanism or action is not known (1). At present the classical picture of kernicterus in surviving newborns has practically disappeared. However, there may be other forms of bilirubin encephalopathy (2). The relationship between neonatal hyperbilirubinemia and neurodevelopmental outcome is controversial. The results of a large national survey from the Netherlands (3) have shown that there is a correlation between maximum serum total bilirubin concentration in the neonatal period and cerebral palsy (CP) in preterm infants. In another recent study from the United States (4) there was no correlation between maximum bilirubin levels and CP or other neurodevelopmental disorders in infants with a birth weight less than 2000 g. A relationship between severe hypocarbia during the first days of life and later brain damage has also been suggested ( 5 ) . Periventricular leukomalacia (PVL), on the other hand, has been shown to be the most important predictor of neurologic handicap in preterm infants (6, 7). In this study we present evidence that both hypocarbia and bilirubin toxicity may have a role in the pathogenesis of PVL.

Patients and methods Between May 1, 1984 and May 31, 1987, 11 869 live infants were born at Tampere University Hospital.

There were 138 infants with a gestational age less than 33 completed weeks. All these infants were investigated in our prospective study of neurodevelopmental outcome in preterm infants. Thirty died during the neonatal period ( < 2 8 days) and five at one to three months of age. According to our routine practice, serum total bilirubin level was determined in all premature infants by micromethod ( A 0 Bilirubinometer, American Optical Co, Buffalo, NY) at 08:OO and 15:OO during the first five days, at 08:OO during the next seven days, and thereafter as required. Phototherapy and exchange transfusions were performed according to the guidelines presented by Sorto (8), which are in close accordance with those of Pearlman et al. (9). Blood-gas values were estimated by the standard Micro-Astrup method from arterial blood samples, and only in some cases were pH and PC02 values estimated from venous or capillary samples. The interval between blood-gas measurements was usually 4-6 h. The duration of carbon dioxide tension (PC02) < 30 mmHg during the first 72 h was calculated from the consecutive PC02 values. Serial ultrasound studies through the anterior fontanelle with a 5-MHz sector scanner were performed in all (exept two) cases. The first investigation was performed as soon as possible after birth and repeated in most cases at intervals of 1 4 days and later 1-2 weeks until discharge from hospital, and thereafter usually up to six months of age. In two cases ultrasound investigations were not performed properly; none had CP at two years of age.

Relationship of P V L and cerebral palsy

ACTA PEDIATR 81 (1992)

Peri-intraventricular hemorrhage (PIVH) was graded according to Papile et al. (10). Enlargement of the lateral ventricles was classified as mild if the diameter of the ventricle measured in parasagittal section above the glomus of the choroid plexus was 0.5> 1.0 cm, as moderate if it was 1 .O > 1.5 cm, and severe if it was

=- 1.5

cm. PVL was classified as mild when persistent periventricular echodensity with small cysts (diameter of 2-3 mm) was found, and extensive when there were large cysts leadng to moderate enlargement of the lateral ventricles. Periventricular echogenicity without cyst formation and solitary periventricular or germinal matrix cysts were classified as normal variants. The treatment practices and other laboratory meth-

ods used in our neonatal unit, determination of gestational age, and follow-up after discharge from hospital were as presented earlier (1 1). All these measures are part of our normal care of preterm infants and were accepted by the parents. At two years of age a neurologic assessment was performed. As the follow-up was relatively short, only clear handicaps were noted. These were CP, which was subdivided into hemiplegia (arm and leg of one side affected), diplegia (legs more severely affected than arms), and tetraplegia (all four limbs affected, arms more severely than legs), mental retardation (IQ < 70), and visual disorders (severe handicaps such as retrolental fibroplasia). According to the neurologic assessment

Tuble 1 . Different perinatal variables analyzed in the study.

Mother’s age Mother’s diseases Diabetes Hypertension Pre-eclampsia Urinary tract infection Pregnancy Antepartum hemorrhage Premature contractions Mother’s medication during delivery Analgesics Anesthesia Antihypertensives p-sympatomimetics Corticosteroids Sedatives Oxytocin Delivery Rupture of membranes Fetal distress Mode of delivery Infant’s condition at birth Apgar score at 1 min Apgar score at 5 min Birth weight Gestational age Fetal distress Resuscitation Sex Infant’s diseases Apnea episodes Bronchopulmonary dysplasia (BPD) Convulsions Hypoglycemia Hypernatremia Hyperviscosity Hyponatremia Necrotizing enterocolitis (NEC) Patent ductus arteriosus (PDA) Pneumothorax Respiratory distress syndrome (RDS) Respiratory disorder (other) Septicemia

803

Laboratory parameters during the first 72 h Highest Pa02 Lowest Pa02 Highest PC02 Lowest PC02 Highest pH Lowest pH Lowest BE Highest systolic blood pressure Lowest systolic blood pressure Duration of PC02 Q 30 mmHg Duration of PCO2 > 60 mmHg Total fluids administered Other variables during the first six days First capillary hematocrit value Capillary hematocrit value (days 1-6) Highest serum total bilirubin value Serum total bilirubin values (days 1-6) Age at highest serum total bilirubin Serum albumin Bilirubin/albumin ratio Lowest serum calcium Lowest thrombocytes Treatment procedures during the first months Surgical operations Duration of oxygen therapy ( > 40%) Duration of phototherapy Duration of respirator treatment Duration of theophylline treatment Number of exchange transfusions

804

RS Ikonen el al.

the series was divided into three groups: infants without CP (n=83), infants with diplegia (n= lo), and infants with quadriplegia (n= 10). There were no cases of hemiplegia at follow-up. One baby without CP had retrolental fibroplasia. According to the study scheme, 64 different variables comprising perinatal data, neonatal morbidity and various laboratory parameters were collected and evaluated (Table 1). In order to screen possible risk factors related to PIVH, PVL and CP the data were first analyzed using Fisher’s exact test or chi-square test for categorical variables, and Student’s two-sided t-test or analysis of variance for continuous variables (1 2). These risk factors were then used as independent variables in logistic regression models (12, 13) on which the final conclusions were based. The logistic regression models were selected using a stepwise technique. The fit of the final model was measured using goodness-of-fit chisquare where a high p value indicates a satisfactory model. Odds ratio (OR) with confidence intervals were applied where appropriate. Because of our basic set-up, the independent variables with negative regression coefficients were associated with CP (i.e. negative outcome) and ORs were below 1. The analysis was carried out on a VAX/VMS computer using BMDP statistical software.

ACTA PEDIATR 81 (1992)

Table 2. Sex, mean ( fSD) birth weight and gestationalage in the three groups of infants.

CP syndrome N o CP (n=83)

Diplegia Tetraplegia (n=lO) (n=10) pvalue

515 317 Sex (female/male) 43/40 Birth weight (9) 1450+370 1440k260 1430k530 Gestational age (weeks) 30.1 1.6 29.0+ 1.7 29.9k 2.4

+

a

0.48‘ 0.99b 0.76b

Chi-square test, 2 x 2 table: no CP versus CP syndrome; ANOVA.

Table 3. Blood-gas values (mean SD) during the first 72 h of life.

CP syndrome N o CP (n = 63)” 7.17k 0.10 Lowest pH Highest pH 7.47 f 0.08 Lowest Pa02b 4 2 5 14 Lowest PCOzb 24 k6.9 Highest PC02 58+18 Duration of PC02 20 17 < 30 mmHg (h)

+

a

Diplegia (n = 8)’ 7.10f0.19 7.52k0.08 40+ I3 20k4.5 59k 19 34k8.3

Tetraplegia (n = 10) p value

-

7.21 k0.08 7.50+0.06 47+ I4 20 6.7 52k 10 35+ 13

+

0.08 0.17 0.56 0.14 0.61 0.004

Because of good condition not monitored in all cases; mmHg.

Results There were no differences between the groups according to sex, birth weight and gestational age (Table 2). There were also no significant differences in other perinatal data or neonatal morbidity (variables presented in Table I). Blood-gas values during the first 72 h are presented in Table 3. The mean duration of PCO2 < 30 mmHg was significantly correlated with the severity of neurologic handicap (p = 0.004). The mean serum total bilirubin levels during the first days of life are shown in Table 4. There was no significant correlation between mean bilirubin values during the first days or the maximum serum bilirubin concentration in the neonatal period and CP. There was, however, a trend for higher bilirubin values in the group of tetraplegic infants compared to the other groups. This difference was statistically significant at the age of three days (t-test, p = 0.044). No significant differences were found for bilirubin/albumin ratio between the groups. There were also no differences between the groups in monitored systolic blood pressure values during the first 72 h. The number of exchange transfusions and the length of phototherapy did not differ between the groups. Table 5 shows the duration of PC02

Hyperbilirubinemia, hypocarbia and periventricular leukomalacia in preterm infants: relationship to cerebral palsy.

This study comprised 103 preterm infants with a gestational age less than 33 weeks who were born in Tampere University Hospital and who were followed ...
544KB Sizes 0 Downloads 0 Views