Intraventricular hemorrhage and cystic periventricular leukomalacia are severe complications of prematurity, and they are the main antecedents of cerebral palsy and neurodevelopmental delay in preterm infants.1 They can present independently or be associated with each other, and they occur most frequently in neonates born at less than 34 weeks of gestation. When present in their higher grades, both lesions may indicate the presence of serious brain injury with risk of neurodevelopmental delay in survivors.1,2 Although release of cytokines leading to disturbance of fetal perfusion and damage from oxygen-derived free radicals are considered the final common pathways in both intraventricular hemorrhage and periventricular leukomalacia,2,3 the initiating factors and processes are the object of debate. Several studies have examined postnatal risk factors associated with these 2 conditions, and risk factors have been identified only in a minority of cases with intraventricular hemorrhage and periventricular leukomalacia.2,4,5 These results suggest that the original insult in both conditions most likely occurs in utero.6,7
An analysis of the studies on the antenatal risk factors for intraventricular hemorrhage and periventricular leukomalacia shows intriguing similarities between these 2 conditions. Acute placental inflammation is the strongest independent prenatal predictor of early-onset neonatal intraventricular hemorrhage in spontaneous prematurity.5 Similarly, a meta-analysis showed that both clinical and histologic chorioamnionitis are independent predictors for the development of cystic periventricular leukomalacia.8 The purpose of our study was to evaluate the clinical and histopathologic factors associated with neonatal intraventricular hemorrhage and periventricular leukomalacia in a cohort of very preterm infants and to establish whether the 2 conditions have different risk factors.
MATERIALS AND METHODS
We accessed a database of singleton neonates born at 24 to 33 weeks of gestation at our institution between January 1,1993 and December 31, 2002. San Gerardo Hospital is a tertiary care center that accepts referrals from a geographic area comprising 9,000 deliveries. The obstetric and neonatal care did not change significantly during the study period (data available from the authors upon request), with the exception of high-frequency oscillatory ventilation, which was introduced in 1997. Excluded from the analysis were neonates transferred from outside hospitals (n = 116), stillbirths, and neonates with congenital anomalies. The database pertaining to deliveries before 34 weeks of gestation is updated on a monthly basis; pertinent information is entered by an assigned physician, and it is reviewed periodically by a Senior Consultant. The clinical charts and radiologic reports of the included neonates were reviewed to identify cases that developed intraventricular hemorrhage or periventricular leukomalacia or both.
Neonatal cranial ultrasound was performed in every neonate, using a standardized protocol for screening of cerebral damage, with exams were performed at 1 day, 3 days, and 7 days of life, and repeated weekly or more frequently as clinically indicated until 40 postconceptional weeks. All scans were done and interpreted by the same team of operators. Standard coronal and parasagittal images were obtained (Acuson XP10 and Acuson Aspen, Acuson Corporation, Mountain View, CA) using a 7-MHz transducer through the anterior fontanel. For the purpose of the study, we included only lesions classified as intraventricular hemorrhage or periventricular leukomalacia at 30 days of life. Intraventricular hemorrhage was diagnosed in the presence of hyperechogenicity in the lateral ventricles, and was classified in 4 grades according to Volpe et al9 Intraventricular hemorrhage of any grade was included in this analysis. Periventricular leukomalacia was defined as periventricular echodense lesions persisting for more than 15 days, evolving or not into cystic lesions; or cystic lesions in the absence of a prior echodense lesion. The lesions were classified in 3 grades according to De Vries classification,10 and all 3 grades were included in our study. Excellent correlation of such ultrasound findings with neuropathologic data has been reported.11 In addition, symmetrical or asymmetrical ventricular dilatation, in the absence of hydrocephaly or intraventricular hemorrhage at a cranial ultrasonographic examination up to 40 postmenstrual weeks is suggestive of white matter disease and therefore was included in the periventricular leukomalacia group.11 Cases with both intraventricular hemorrhage and periventricular leukomalacia were assigned to the group more representative of the dominant lesions (eg, a case with intraventricular hemorrhage grade 1 and periventricular leukomalacia grade 2 would be assigned to the periventricular leukomalacia group). Neurologic outcome was assessed every 3 months for the first year, at 18 months, and yearly thereafter using the Milani-Comparetti and Gidoni scoring system.12
Gestational age was based on the date of the last menstrual period and prenatal ultrasound examination performed before 22 weeks of gestation, which is routinely done at our institution. Small for gestational age was defined as a birth weight less than 10th centile for sex and gestational age according to Italian standards. Clinical chorioamnionitis was diagnosed in the presence of 2 or more of the following: temperature elevation of at least 38°C, white blood cell count at least 15,000 cells per millimeter, uterine tenderness, and foul-smelling vaginal discharge. Neonatal sepsis was diagnosed in the presence of positive blood cultures. Administration of steroids to enhance fetal lung maturity was considered for our analysis only if 48 hours elapsed from the initial dose of steroids.
Histopathologic examination of the placenta was performed by the same observer, who was blinded to the neonatal status, and the lesions were classified according to standard published protocols,13 with placental information available in 567 of the 653 (86.8%) infants of our final sample. Acute inflammatory lesions and uteroplacental vascular lesions were classified as present or absent independently from the severity or extent of the lesions. The study was approved by the Institutional Review Board committee of San Gerardo Hospital, Monza, Italy.
Clinical and histopathologic data were compared between cases with any cerebral lesion and those without lesions and between those with intraventricular hemorrhage and periventricular leukomalacia using Student t test for continuous variables and χ2 or Fisher exact test for categorical variables. Stepwise logistic regression analysis was conducted to identify the independent antenatal predictors of intracranial ultrasound lesions. Receiver operating characteristic curve analysis was used to assess the sensitivity and specificity of prenatal variables that are predictors of postnatal cerebral lesions. A 2-tailed P < .05 was considered significant.
A total of 653 neonates born between 24 weeks and 33 weeks of gestation fulfilled the study criteria, including 207 born after preterm rupture of membranes, 145 after spontaneous preterm labor, and 301 after indicated preterm delivery. A diagnosis of ultrasonographic cerebral lesion was made in 64 neonates at a median of 6 (range 1–12) neonatal cranial ultrasound examinations. Intraventricular hemorrhage was diagnosed in 45 neonates and periventricular leukomalacia in 19. Both intraventricular hemorrhage and periventricular leukomalacia were present in 9 of 64 cases (14%). When we compared the demographic and antenatal variables between neonates with any type of cerebral lesion and the remaining neonates in the cohort (Table 1), we noted significantly higher rates of spontaneous prematurity, premature rupture of membranes, clinical chorioamnionitis, and use of antibiotic therapy in neonates with cerebral lesions. Neonates destined to develop intraventricular hemorrhage or periventricular leukomalacia were born at an earlier gestational age and weighed less than those who did not develop either sonographic cerebral lesion (Table 2). Of interest, the umbilical artery pH was similar between the 2 groups, thus suggesting that peripartum hypoxia was not a main causative process of intraventricular hemorrhage and periventricular leukomalacia. At placental histopathologic examination, the group with cerebral lesions had similar rates of uteroplacental vascular damage as controls (34.9% or 22 of 63 compared with 43.0% or 217 of 504, P = .27), but higher rates of acute inflammatory lesions (46.0% or 29 of 63 compared with 24.8% or 125 of 504, P = .001).
When all the prenatal predictors of cerebral lesions found to be significant or approaching significance at univariate analysis (ie, P < .10) were entered into a forward conditional stepwise logistic regression analysis (including spontaneous preterm labor, antibiotics, antenatal steroids, preeclampsia, chorioamnionitis, premature rupture of membranes, labor before delivery, gestational age at delivery, birth weight, gender, Pco2, Po2, and placental lesions), only spontaneous prematurity (odds ratio = 1.9; 95% confidence interval [CI] 1.1–3.4) and gestational age at delivery (odds ratio = 0.8; 95% CI 0.7–0.9) were independently associated with odds of intraventricular hemorrhage and periventricular leukomalacia. Figure 1 displays the probability of development of cerebral lesions according to gestational age at delivery and type of prematurity (spontaneous or indicated). Receiver operating characteristic curve analysis applied to the model produced by the stepwise regression analysis (containing only type of prematurity and gestational age at delivery) demonstrated that assuming a cutoff that equally balances sensitivity and false positive rate, the sensitivity would be 67% (Fig. 2). In other words, two thirds of the cases of cerebral lesions can be predicted on the basis of spontaneous prematurity and gestational age at delivery alone, implying that only the remaining one third need be attributed to other prenatal or postnatal causes.
We then compared the prenatal characteristics within the group of cerebral lesions between the 2 groups of infants with intraventricular hemorrhage or periventricular leukomalacia. The 2 groups did not differ in any of the antenatal variables (Tables 3 and 4). Histopathologic placental findings were nonsignificantly different between the periventricular leukomalacia and intraventricular hemorrhage groups (uteroplacental vascular lesions, 5 of 19 or 26% compared with 17 of 44 or 39%, respectively, P = .51; acute inflammatory lesions, 6 of 19 or 31% compared with 23 of 44 or 52%, respectively, P = .22). The observed difference of 21% in acute inflammatory lesions (ie, 31% compared with 52%) had a 95% CI for the difference ranging from –10% to +45%. Our study had adequate power to detect a difference of 33% between the 2 groups in inflammatory placental lesions (alpha = 0.05; β = 0.80). In other words, 161 cases would be needed in each group for the observed difference to reach statistical significance.
Neonatal death occurred in 2 cases with periventricular leukomalacia and 5 with intraventricular hemorrhage. Five cases were lost to follow-up. The rate of neurodevelopmental delay at a mean follow-up of 30.7 ± 29.7 months was higher in neonates with periventricular leukomalacia (13 of 15 or 86.7%) compared with intraventricular hemorrhage (16 of 37 or 43.2%, P = .01).
The pathophysiologic mechanism of intraventricular hemorrhage and periventricular leukomalacia is thought to involve release of cytokines leading to disturbance of fetal perfusion, with acute fluctuations in cerebral blood flow in the preterm fetus and neonate, where autoregulation is impaired.2,4,5,14 According to this hypothesis, a predominant increase in cerebral blood flow would predispose to intraventricular hemorrhage, whereas cerebral ischemia would predispose to periventricular leukomalacia. The initiating factors and processes most likely occur in utero. However, when we compared neonates with intraventricular hemorrhage with those with periventricular leukomalacia, we could not identify any significant difference in prenatal variables. This could be due to limited power of our study, or it may suggest that both conditions share similar prenatal causative factors. Indeed, intraventricular hemorrhage and periventricular leukomalacia often coexist, as occurred in 14% of cases in our study. Experimental studies in animals suggest that differences in gestational age at the time of in utero insult, duration of the insult, or degree of severity of the insult may explain the different clinical manifestations.3
We have found that nearly two thirds of cases of intraventricular hemorrhage and periventricular leukomalacia can be predicted by spontaneous prematurity and gestational age at delivery, ie, have causative processes that occur in the prenatal period. Infection seems to be the most important causative process involved in both intraventricular hemorrhage and periventricular leukomalacia, as suggested by the significant association of clinical and histologic chorioamnionitis with these lesions at univariate analysis. Other studies have made the same observation in populations with only intraventricular hemorrhage5,15 or periventricular leukomalacia.6,8,16,17 Indirect evidence implicating intrauterine infection in the pathogenesis of intraventricular hemorrhage and periventricular leukomalacia comes from several studies showing positive amniotic fluid cultures and high interleukin-6 levels in the cord blood of affected neonates.6,15,17–21 Fetal infection may lead to intraventricular hemorrhage or periventricular leukomalacia by different pathways. For example, bacterial toxins may severely alter fetal cardiovascular function, resulting in dysregulation of cerebral blood flow; and cytokines released during intrauterine infection may directly cause neuronal damage.22 However, evidence of infection was present in less than one half of our cases with intraventricular hemorrhage or periventricular leukomalacia, suggesting that both types of intracranial lesions are heterogeneous in their causative processes, and that intrauterine infection is only one of the causative factors. Uteroplacental insufficiency is one of the alternative processes that has been proposed to play a causative role in the occurrence of these sonographic cerebral lesions.23 However, in our study none of the common clinical (ie, preeclampsia or small for gestational age neonates) or histologic correlates of uteroplacental insufficiency was significantly associated with intracranial cerebral lesions.
Our multivariate analysis shows that clinical and histologic indicators of intrauterine infection lose statistical significance after controlling for spontaneous prematurity and gestational age at delivery. Neonates born after preterm labor or premature rupture of membranes are twice as likely to develop intraventricular hemorrhage and periventricular leukomalacia as those delivered for maternal or fetal reasons, as already reported.18 Moreover, there is an inverse correlation between gestational age at delivery and risk of cerebral lesions both in the neonates delivered spontaneously and for maternal or fetal indications. This observation suggests that a vulnerability may be present in the fetal brain at low gestational ages irrespective of the pathologic process that causes prematurity itself. It is also possible that more severe intrauterine insult leads both to lower gestational age at delivery and more severe intracranial lesions. We have provided risk figures that can assist the clinician in the identification of the prenatal risk profile for neonatal cerebral lesions. For example, a patient presenting in preterm labor or premature rupture of membranes with impending preterm delivery at 24 weeks of gestation has a 30% risk of intraventricular hemorrhage or periventricular leukomalacia, which carries a greater than 50% risk of subsequent neurodevelopmental disability. Prophylaxis with antibiotics and steroids, as is currently recommended, does not seem to reduce the risk of cerebral lesions significantly. As a corollary, our study results provide further support to the evidence that mode of delivery (vaginal or cesarean) does not have an independent effect on the occurrence of cerebral lesions after controlling for indicators of intrauterine or fetal infection.5,24,25
In summary, the emerging picture is that antenatal or intrapartum insult triggers are primarily responsible for the causation of intraventricular hemorrhage or periventricular leukomalacia. Intrauterine infection is one of the critical factors identified in the causation of these lesions. Neonates born after spontaneous prematurity and at low gestational ages are at highest risk for intraventricular hemorrhage and periventricular leukomalacia.
1.Hack MB, Taylor HG. Perinatal brain injury in preterm infants and later neurobehavioral function. JAMA 2000;284:1973–4.
2.Weindling M. Periventricular hemorrhage and periventricular leukomalacia. Br J Obstet Gynaecol 1995;102:278–81.
3.Volpe JJ. Neurobiology of periventricular leukomalacia in the premature infant. Pediatr Res 2001;50:553–62.
4.Funato M, Tamai H, Noma K, Kurita T, Kajimoto Y, Yoshioka Y, et al. Clinical events in association with timing of intraventricular hemorrhage in preterm infants. J Pediatr 1992;121:614–9.
5.Vergani P, Patanè L, Doria P, Borroni C, Cappellini A, Pezzullo JC, et al. Risk factors of neonatal intraventricular haemorrhage in spontaneous prematurity at 32 weeks gestation or less. Placenta 2000;21:402–7.
6.Leviton A, Paneth N. White matter damage in preterm newborns: an epidemiologic perspective. Early Hum Dev 1996;44:1–16.
7.Trounce JQ, Shaw DE, Levene MI, Rutter N. Clinical risk factors and periventricular leukomalacia. Arch Dis Child 1988;63:17–22.
8.Wu YW, Colford JM. Chorioamnionitis as a risk factor for cerebral palsy: a meta-analysis. JAMA 2000;284:1417–24.
9.Volpe JJ. Neurology of the newborn. 4th ed. Philadelphia (PA): W.B. Saunders; 2001.
10.De Vries LS, Eken P, Dubowitz LMS. The spectrum of leukomalacia using cranial ultrasound. Behav Brain Res 1992;49:1–6.
11.Ment LR, Bada HS, Barnes P, Grant PE, Hirtz D, Papile LA, et al. Practice parameter: neuroimaging of the neonate: report of the Quality Standards Subcommittee of the American Academy of Neurology and the Practice Committee of the Child Neurology Society. Neurology 2002;58:1726–38.
12.Milani-Comparetti A, Gidoni EA. Routine developmental examination in normal and retarded children. Dev Med Child Neurol 1967;9:631–6.
13.Lowell DM, Kaplan C, Salafia CM. College of American Pathologists Conference XIX on the examination of the placenta: report of the working group on the definition of structural changes associated with abnormal functions in the maternal-fetal-placental unit in the second and third trimesters. Arch Pathol Lab Med 1991;115:647–731.
14.Lou HC. The “lost autoregulation hypothesis” and brain lesions in the newborn—an update. Brain Dev 1998;10:143–6.
15.Developmental Epidemiology Network Investigators. The correlation between placental pathology and intraventricular hemorrhage in the preterm infant. Pediatr Res 1998;43:15–9.
16.Perlman J, Risser R, Broyles S. Bilateral cystic periventricular leukomalacia in the premature infant: associated risk factors. Pediatrics 1996;97:822–7.
17.Damman O, Leviton A. Maternal intrauterine infection, cytokines, and brain damage in the preterm newborn. Pediatr Res 1997;42:1–8.
18.Verma U, Tejani N, Klein S, Reale MR, Beneck D, Figueroa R, et al. Obstetric antecedents of intraventricular hemorrhage and periventricular leukomalacia in the low-birth-weight neonate. Am J Obstet Gynecol 1997;176:275–81.
19.Gomez R, Romero R, Ghezzi F, Yoon BH, Mazor M, Berry SM. The fetal inflammatory response syndrome. Am J Obstet Gynecol 1998;179:194–202.
20.Salafia CM, Minior VK, Rosenkrantz TS, Pezzullo JC, Popek EJ, Cusick W, et al. Maternal placental and neonatal associations with early germinal matrix/intraventricular hemorrhage in infants born before 32 weeks’ gestation. Am J Perinatol, 1995;12:429–36.
21.Weeks JW, Reynolds L, Taylor D, Lewis J, Wan T, Gall SA. Umbilical cord blood interleukin-6 levels and neonatal morbidity. Obstet Gynecol 1997;90:815–8.
22.Berger R, Garnier Y, Jensen A. Perinatal brain damage: underlying mechanisms and neuroprotective strategies. J Soc Gynecol Investig 2002;9:319–28.
23.Van Gelder-Hasker MR, Van Wezel-Meijler G, De Groot L, Van Geijn HP, De Vries JIP. Peri- and intraventricular cerebral sonography in second- and third-trimester high risk fetuses: a comparison with neonatal ultrasound and relation to neurological development. Ultrasound Obstet Gynecol 2003;22:110–20.
24.Spinillo A, Capuzzo E, Stronati M, Ometto A, De Santolo A, Acciano S. Obstetric risk factor for periventricular leukomalacia among preterm infants. Br J Obstet Gynaecol 1998;105:865–71.
25.Hansen A, Leviton A. Labor and delivery characteristics and risk of cranial ultrasonographic abnormalities among very-low-birth-weight infants. The Developmental Epidemiology Network Investigators. Am J Obstet Gynecol 1999;181:997–1006
© 2004 by The American College of Obstetricians and Gynecologists. Published by Wolters Kluwer Health, Inc. All rights reserved.
This article has been cited