Cranial Ultrasound Abnormalities Identified at Birth: Their Relationship to Perinatal Risk and Neurobehavioral Outcome
Objectives. Minor cranial ultrasound abnormalities, such as mild ventricular enlargement, choroid plexus cysts, and subependymal cysts, have been identified in 3% to 5% of the newborn population. Although clinicians generally consider these abnormalities to be insignificant for the outcome of the newborn, few convincing data have been published to support this optimism. The objectives of this study were to identify potential risk factors associated with the identification of cranial ultrasound abnormalities at birth and to determine if the abnormalities were related to neurobehavioral sequelae in the newborn.
Methods. Three hundred eight women were enrolled in this prospective, longitudinal maternal-infant health and development study either at the time they entered the public health care system for prenatal care or at delivery if they had no prenatal care. Each woman participated in an in-depth psychosocial interview at the end of each trimester of pregnancy. Retrospective chart review by experienced medical personnel was used to compile data for the Hobel perinatal risk score for each study participant after delivery. Offspring underwent cranial ultrasound evaluation, the Amiel-Tison Neurologic Assessment, and the Brazelton Neonatal Behavioral Assessment Scale within 96 hours of birth by experienced examiners blinded to any maternal-infant history.
Results. Of the 308 women originally enrolled in the study, 301 delivered living infants. Of these, 266 infants (88%) underwent a cranial ultrasound evaluation and are the subject of this article. For the purposes of the current study, infants were divided into those with normal (n = 239) and those with abnormal (n = 27) ultrasound results. Abnormal ultrasound results included the following lesions: subependymal cyst (n = 13); mild ventricular enlargement (n = 6); choroid plexus cysts (n = 3); a combination of cysts and increased ventricular size (n = 2); a 7-mm midline cyst in the superior posterior portion of the third ventricle (n = 1); subependymal hemorrhage and ventricular enlargement (n = 1); and increased ventricular size, subependymal hemorrhage and cysts, and two small, right thalamic calcifications (n = 1). There were no significant differences between those with an abnormal ultrasound and those with a normal ultrasound for birth weight, length, gestational age, rate of prematurity, frequency of nulliparity, or frequency of small for gestational age infants. However, infants with an abnormal ultrasound had a significantly smaller mean head circumference than those with a normal ultrasound (34.5 ± 1.9 cm vs 33.7 ± 1.9 cm). The infants with an abnormal ultrasound had a higher median prenatal (50 vs 45), neonatal (14 vs 8), and total (94 vs 77) Hobel risk score but not a higher labor-delivery score. There were no significant differences when these groups were compared on additional risk factors not included in the Hobel scoring system such as race and socioeconomic status. In addition, mothers who used a greater number of drugs during the first trimester of pregnancy were more likely to have an infant with an abnormal ultrasound at birth such that the probability of having an abnormal ultrasound rose to 22% by the time the pregnant women were using four drugs. Neurologic examinations revealed no differences between the infants with normal and abnormal ultrasounds. There were also no group differences for five of the seven Brazelton cluster scores, the excitable or depressed clusters, or eight of the nine qualifier scores. However, infants with abnormal ultrasounds performed significantly better on the habituation (7.3 ± 0.8 vs 6.6 ± 1.5) and autonomic regulation (6.5 ± 0.8 vs 6.0 ± 1.0) clusters but more poorly on the cost of attention qualifier score (4.9 ± 1.2 vs 5.5 ± 1.2) on the Brazelton Neonatal Behavioral Assessment Scale.
Conclusion. Infants with an abnormal cranial ultrasound at birth had higher perinatal risk scores. Additionally, we conclude that the abnormalities identified on cranial ultrasound within the first 96 hours of life were most likely benign with no clinical significance in the immediate neonatal period. However, because of the finding of several subtle differences in behavior, follow-up beyond the neonatal period would be advisable to determine if any late-onset abnormalities occur.
With the advent of ultrasound technology in the early 1980s, physicians acquired a rapid, noninvasive, bedside opportunity to investigate the anatomy of the newborn brain. Previous attempts to view the newborn brain required either exposure to radiation with a computed tomographic scan or could not be done until autopsy. As with any new technology there followed a rapid increase in knowledge regarding normal variants of anatomy of the newborn brain as well as a growing familiarity with potential abnormalities, including such problems as subependymal cysts, choroid plexus cysts, and mild ventricular enlargement.
Previous studies, including our own high-risk sample of drug-exposed newborns have attempted to investigate the etiology and significance of these lesions. However, little data exist regarding a systematic, detailed evaluation of risk factors associated with the development of these ultrasound abnormalities or regarding the neurobehavioral outcome of infants so afflicted. Thus the present investigation was undertaken with the following objectives: 1) to identify potential risk-factors associated with ultrasound abnormalities identified at birth; and 2) to determine if the abnormalities identified were related to neurobehavioral abnormalities in the immediate newborn period. The hypotheses were that 1) infants determined to be at higher perinatal risk would be more likely to have cranial ultrasound abnormalities identified at birth; and 2) infants with cranial ultrasound abnormalities identified at birth would perform more poorly on standardized tests of newborn neurobehavioral function.
This study was approved by the University of Florida Health Science Center Institutional Review Board. Informed consent was obtained from all study participants. Confidentiality of data was maintained under Confidentiality Certificate No. DA-91–45 from the National Institute on Drug Abuse, United States Department of Health and Human Services.
This study sample was drawn from a cohort of children participating in a prospective growth and development study designed to elucidate the effects of prenatal drug exposure on the developing fetus and child. A full description of participant selection and study design for the parent study has been previously published.1 In summary, after meeting a priori exclusion and inclusion criteria, 154 prenatal cocaine users were enrolled at the time they presented for prenatal care or at delivery, in the case of limited or no prenatal care, and matched on race, parity, socioeconomic status, and prenatal risk to 154 noncocaine users for a total enrollment of 308 women. The women in the study participated in a detailed interview at the end of each trimester of pregnancy. The interview included questions about previous and current obstetrical problems, psychosocial and lifestyle measures, and detailed questions regarding amount, frequency, and patterns of licit and illicit drug use. Drug exposure was further determined by the results of two urine specimens, one obtained at study entry and one at delivery, screened for barbiturates, benzodiazapines, methamphetamines/amphetamines, propoxyphene, phencyclidine, opiates, cocaine, and cannabinoid metabolites by fluorescence polarization-immunoassay methods2 with positive screens confirmed by gas chromatography/mass spectrometry. Study participants were eligible for inclusion in the current report if their cranial ultrasound and neurobehavioral assessments were performed within 96 hours of birth.
Perinatal risk was determined by experienced medical personnel using the well-published scoring system of Hobel et al.3Information for the Hobel was obtained by retrospective chart review after the birth of the infant. The score is divided into three components, prenatal, labor-delivery, and neonatal, as well as summed for a total risk score. The prenatal component consists of 30 questions regarding historical problems that developed before the current pregnancy and 31 questions about historical problems that developed during the current pregnancy. For example, problems that developed before the current pregnancy include questions about previous pregnancy outcomes, family history, and maternal medical-surgical problems. Problems that developed during the current pregnancy include questions about maternal serology, infection status, bleeding, and hypertension. The labor-delivery component is divided into 17 questions about early labor, 11 about interim labor, and 15 about problems that developed during late labor. Finally, the neonatal component consists of 69 questions that cover anything that occurs while the infant is hospitalized after delivery.
All infant assessments were performed within the first 4 days of life by assessors blinded to maternal or infant history.
Offspring underwent cranial ultrasound examination by an experienced technologist using a 7.5-mHz transducer and contemporary equipment. Standard coronal and parasagittal images were obtained through the anterior fontanelle. After the completion of all study births, all ultrasounds were read by one experienced pediatric radiologist (W.C.), who reviewed each ultrasound in a standardized manner. This included evaluation for anatomy, maturity, ventricular size, calcifications, hemorrhages, cyst formation, and leukomalacia. Ventricular size was measured in the anterior horns on coronal images through Monro's foramina. Because fewer than 5% of infants will have a measurement >2 mm, those >3 mm were considered to be enlarged.
The Brazelton Neonatal Behavioral Assessment Scale4 was performed by experienced, certified developmental examiners in the Clinical Research Center under controlled conditions of light, temperature, and sound, and approximately midway between feedings. Because the Brazelton Neonatal Behavioral Assessment Scale does not provide one summary score, individual scores were summarized using the seven behavioral clusters established by Lester et al5: habituation, orientation, motor maturity, state lability, state regulation, autonomic regulation, and abnormal reflexes, as well as the recent excitable/depressed cluster scheme.6 In addition, the nine supplementary scores used to qualify the infant's behavioral response by evaluating the effort and physiologic cost of an infant's performance were assessed.
The Amiel-Tison Neurologic Assessment7 was performed by experienced neonatal nurse practitioners in the Clinical Research Center. Individual items from the assessment were grouped into the following domains: ocular signs, sensory development, posture and spontaneous motor activity, passive tone, active tone, primary reflexes, and deep tendon reflexes. Each item was recorded by the examiner per the standardized examination score sheet. For any individual infant the number of abnormal findings within each domain were summed and divided by the number of items in that domain. These scores, which ranged between 0 and 1 for each of the seven domains, were used for statistical analyses.
Study infants were grouped according to whether or not their cranial ultrasound at birth was normal or abnormal. Continuous data were analyzed using either mean values and Student's t test or median values and the Wilcoxon rank sum test for comparisons. Comparisons of categorical data were made using a one-tailed Fisher's exact test. Logistic regression modeling was used to determine the relationship between the number of drugs a women used during pregnancy (0–4 drugs) and an abnormal ultrasound result. The statistical significance level was set at P < .05.
Three hundred eight women were enrolled in the parent study. Forty-one percent were enrolled at the end of the first trimester, 34% at the end of the second trimester, and 25% at delivery. These women gave birth to 301 living offspring. Of these, 266 infants (88%) underwent a cranial ultrasound evaluation and are the subject of this article. A full description of this subsample, including the types of abnormalities observed, and their relationship to prenatal cocaine exposure, has been reported elsewhere.8
For purposes of the current study, infants were divided into those with normal (n = 239) and those with abnormal (n = 27) ultrasound results. Abnormal ultrasound results included the following lesions: subependymal cysts (n = 13); ventricular enlargement (n = 6); choroid plexus cysts (n = 3); cysts and ventricular enlargement (n = 2); a 7-mm midline cyst in the superior posterior portion of the third ventricle (n = 1); subependymal hemorrhage and ventricular enlargement (n = 1); and subependymal hemorrhage, ventricular enlargement, subependymal cysts, and two small, right thalamic calcifications (n = 1).
There were no significant differences between those with an abnormal ultrasound and those with a normal ultrasound for birth weight (3095 ± 921 g vs 3111 ± 639 g; P= .9297), length (49.8 ± 3.5 cm vs 49.1 ± 3.4 cm;P = .2933), gestational age (37.6 ± 4.2 weeks vs 38.8 ± 2.1 weeks; P = .1516), rate of prematurity (18.5% vs 11.7%; P = .230), frequency of nulliparity (18.5% vs 11.7%; P = .230), or frequency of small for gestational age infants (3.7% vs 6.4%; P = .494). However, infants with an abnormal ultrasound had a significantly larger mean head circumference than infants with a normal ultrasound (34.5 ± 1.9 cm vs 33.7 ± 1.9 cm; P = .0462).
Comparisons of the normal and abnormal ultrasound groups on the Hobel risk scale revealed no significant differences for the labor-delivery subscale. However, infants with abnormal ultrasounds had significantly higher median scores on the prenatal (50 vs 45; P = .0434) and neonatal (14 vs 8; P = .0456) subscales and on the total (94 vs 77; P = .0099) score than those with normal ultrasounds. There were no significant differences when these groups were compared on additional risk factors not included in the Hobel scoring system such as race and socioeconomic status (Table 1).
Analyses of various types and amounts of prenatal drug exposure and exposure patterns were performed. Initial comparisons were made between the normal and abnormal ultrasound groups for use of cigarettes, alcohol, marijuana, and cocaine during the first, second, and third trimester and for the total pregnancy. No significant group differences were found for cigarette, alcohol, or cocaine use during any trimester of pregnancy or for the total pregnancy or for marijuana use during the second and third trimester of pregnancy. However, there were significantly more marijuana users in the abnormal ultrasound group than in the normal group for use during the first trimester (37.0% vs 17.6%; P = .02) and for the total pregnancy (40.7% vs 23.4%; P = .046).
When we analyzed the actual amount of drug use during the various trimesters and the total pregnancy we found no significant group differences for any of the four drugs analyzed (Table 1). A final analysis looked at the effect of the number of different drugs used during the various trimesters of pregnancy and anytime during pregnancy. Abnormal ultrasound outcome was significantly related to an increasing number of drugs used during the first trimester of pregnancy such that the probability of having an abnormal ultrasound rose to 22% by the time the pregnant women were using combinations of four drugs (Table 2).
Two hundred fifty-three infants with an ultrasound (95%) underwent neurobehavioral assessment, including 229 infants with a normal ultrasound and 24 infants with an abnormal ultrasound. There were no significant differences between those with normal and abnormal ultrasounds for five of the seven Brazelton cluster scores, the excitable or depressed clusters, or for eight of the nine qualifier measures. Infants with abnormal ultrasound results did score significantly higher on the habituation (7.3 ± 0.8 vs 6.6 ± 1.5; P = .0074) and autonomic regulation (6.5 ± 0.8 vs 6.0 ± 1.0; P = .0201) clusters and lower on the cost of attention qualifier (4.9 ± 1.2 vs 5.5 ± 1.2;P = .0366) (Table 3).
There were no significant group differences for any of the domains of the neurologic examination (Table 4).
The first objective of this study was to identify any perinatal risk factors related to the identification of cranial ultrasound abnormalities at birth. Using the standardized Hobel risk assessment scale we were able to show that newborns with abnormal cranial ultrasounds had mothers with higher prenatal risk scores but not labor-delivery scores. Thus these infants were born to mothers who were more likely to have historical problems with pregnancy as well as problems developing with the current pregnancy. However, there was not any difference between groups in the development of acute problems during the labor and delivery phase. Again, however, infants at higher risk based on a higher newborn risk score were also more likely to have an abnormal ultrasound at birth. Of particular interest was the association between the number of drugs the infant was exposed to in utero and the increasing risk of having an abnormal ultrasound. Although none of the drugs individually were associated with an abnormal ultrasound, there was something about the sheer number of drugs to which the fetus was exposed that increased the risk for abnormalities to be identified. This study was not designed to determine what the etiology of that association might be. However, we propose that the use of an increasing number of drugs in itself or as a marker for other lifestyle and health variables could affect the development of ultrasound abnormalities in exposed newborns.
The second objective was to determine if the ultrasound abnormalities identified were related to any neurobehavioral abnormalities in the immediate newborn period. The types of ultrasound abnormalities seen in our sample of newborns included choroid plexus cysts, subependymal cysts, ventricular enlargement, and a group of infants with multiple abnormalities. Choroid plexus cysts have been reported in the fetus for more than a decade now; however, their origin remains controversial. Most cysts are transient and resolve before delivery and although some are associated with other fetal malformations, many are not.9–11 Reports of choroid plexus cysts in the newborn are also available. Even less is known about the potential significance of these lesions when they are discovered postnatally. Riebel et al12 concluded that they most likely represented a clinically irrelevant variant of normal. This conclusion was based on data obtained during a 2.5-year period on study infants first scanned between 1 day to 1 year of age because of a variety of risk factors. In their study they obtained general medical and neurologic information from each child's physician or the child's medical record and reported no abnormalities. However, the age at follow-up of the children is not reported and no standardized assessment was used for the neurologic evaluation. These methodologic weaknesses influence the authors' ability to draw valid conclusions from their data.
Subependymal cysts are generally considered to be the result of hemorrhage,13 hypoxic-ischemic damage,14 or neurotropic infection.15 Only rarely are they thought to occur in an otherwise healthy, term newborn.16 Several studies are available that identify subependymal cysts in infants and provide neurologic follow-up of the involved children. For example, Thun-Hohenstein et al16 identified three infants with cysts that appeared on an ultrasound scan obtained after a seizure or apnea spell. These cysts were transient and were related to a diverse clinical picture and a diverse developmental outcome with only one of the three children described as normal. A similar study identified five children with subependymal cysts who underwent ultrasound scanning secondary to clinical abnormalities. Two of the infants died, two were neurologically normal at 6 and 12 months of age, respectively, and one child was clinically abnormal with proven cytomegalovirus infection.15 In a study by Rademaker et al,1724 infants were identified with subependymal cysts from a sample of infants who were ultrasound scanned after admission to the intensive care unit. The cysts seemed to be transient in nature as other investigators have reported. In addition, 22 infants were examined between 3 and 24 months of age with a neurologic examination and the Griffith assessment. Of these 22, all were normal except 2 with transient dystonia and 1 with spastic diplegia. In a study most like the current one, Shen and Huang18 scanned 500 normal neonates and found 25 with subependymal cysts. The obstetric history was normal in 18 of the infants and all were normal neurologically at the time of birth. Again, the ultimate outcome of infants with subependymal cysts remains somewhat controversial. The most reasonable conclusion seems to be that of Mito et al.19 They followed 8 newborns admitted to their intensive care unit and scanned within 24 hours of birth who were found to have subependymal cysts. The children were followed for varying lengths of time ranging from 8 months to >3 years. Three of the children demonstrated abnormalities of the central nervous system. These researchers conclude that infants with subependymal cysts have variable outcomes probably related to the various etiologic pathologies of the cysts.
The issue of increased ventricular size was evaluated by Perry et al20 who evaluated 533 neonates and followed-up 13 of the 15 infants with ventricles >3 mm in width. The evaluations took place at ages varying from 3 to 12 months and although the authors did not indicate how the infants were specifically evaluated, they were all described as neurologically and developmentally normal.
One problem with the available studies on this topic is the inability to gather large sample sizes on which to analyze data and base conclusions. Many of the previous studies performed ultrasound evaluations on large numbers of infants as we did, and also as we did, identified only small numbers of children with abnormalities. What makes our study different is not the sample size but the rigorous, prospective methodology that reduced some of the bias present in previous studies. We performed cranial ultrasounds on a large group of neonates prospectively chosen for study based on a priori inclusion and exclusion criteria that were unrelated to their potential for having an ultrasound abnormality or for being developmentally abnormal. This is in contradistinction to the available studies, most of which chose their sample from a group of newborns and infants referred for cranial ultrasound scanning because of a clinical problem, thereby reducing the ability to generalize results. In addition, all infants were scanned within 96 hours of birth, increasing the likelihood that the abnormalities were perinatal in origin. Several of the previous studies included not only newborns but infants who were first scanned at 1 year of age, complicating interpretation of the age of appearance of the lesions, which may have resulted in varied outcomes. In addition, all of our study infants were evaluated neurobehaviorally under rigorous, standardized conditions at the same age increasing the comparability, reliability, and validity of these data, with all evaluators blinded to any information about the infant, decreasing evaluator bias.
With this rigorous methodology we were able to demonstrate that infants at greater perinatal risk were more likely to have evidence of intracranial abnormalities at birth, particularly if the mother used as many as four different drugs during pregnancy. However, we were unable to identify any clinically significant abnormalities at birth in behavior or neurologic status associated with the identified sonographic abnormalities. These findings lead us to conclude that the abnormalities we identified on ultrasound within the first 96 hours of life, were most likely benign with no clinical significance in the immediate neonatal period. Because of the finding of several subtle differences in behavior, however, follow-up beyond the neonatal period would be advisable to determine if any late-onset abnormalities occur.
This work was supported by National Institutes of Health Grants DA05854 and RR00082.
- Received June 10, 1998.
- Accepted November 10, 1998.
Reprint requests to (M.B.) PO Box 100296, Pediatric Neonatology, University of Florida, Gainesville, FL 32610.
- Eyler FD,
- Behnke M,
- Conlon M,
- Woods NS,
- Wobie K
- ↵Brazelton TB. Neonatal Behavioral Assessment Scale. Clinics in Developmental Medicine (No 88). 2nd ed. Philadelphia, PA: JB Lippincott Co; 1984
- ↵Lester BM, Freier K, LaGasse L. Prenatal cocaine exposure and child outcome: what do we really know? In: Lewis M, Bendersky M, eds. Mothers, Babies, and Cocaine: The Role of Toxins in Development. Hillsdale, NJ: Lawrence Erlbaum; 1995
- ↵Amiel-Tison C, Grenier A. Neurologic Assessment Within the First Year of Life. New York, NY: Oxford University Press; 1986
- Shen E-Y,
- Huang F-Y
- Perry RNW,
- Bowman ED,
- Murton LJ,
- Roy RND,
- de Crespigny LC
- Copyright © 1999 American Academy of Pediatrics