Obstetrics & Gynecology:
Central Nervous System Findings on Fetal Magnetic Resonance Imaging and Outcomes in Children With Spina Bifida
Chao, Tamara T. MD; Dashe, Jodi S. MD; Adams, Richard C. MD; Keefover-Hicks, Amy MSN, RN; McIntire, Donald D. PhD; Twickler, Diane M. MD
From the Departments of Obstetrics and Gynecology, Pediatrics, and Radiology, University of Texas Southwestern Medical Center, Dallas, Texas.
Presented at the Society for Maternal–Fetal Medicine 30th Annual Meeting—The Pregnancy Meeting, February 4, 2010, Chicago, Illinois.
Corresponding author: Tamara T. Chao, MD, Department of Obstetrics & Gynecology, University of Texas Southwestern Medical Center, 5323 Harry Hines Boulevard, Dallas, TX 75390-9032; e-mail: firstname.lastname@example.org.
Financial Disclosure The authors did not report any potential conflicts of interest.
OBJECTIVE: To estimate the relationship between fetal magnetic resonance imaging (MRI) findings of ventriculomegaly, cerebellar herniation, extraaxial space effacement and adverse outcomes in children with spina bifida.
METHODS: This was a review of all pregnancies with spina bifida referred for fetal MRI from 2001 to 2007 and cared for postnatally at our spina bifida referral center. Degree of cerebellar herniation was based on lowest cervical vertebral level reached. Extraaxial space was considered effaced if too small to be measured above or below the tentorium cerebelli. Bladder dysfunction was termed high risk if renal damage was present or urodynamic studies indicated increased risk for renal damage. Ambulation was assessed in children age 3 years or older. Statistical analyses included chi square, Mantel-Haenszel test, and logistic regression.
RESULTS: Magnetic resonance imaging was performed in 36 pregnancies with fetal spina bifida at 27±6 weeks with subsequent delivery at 38±1 weeks. Outcomes were assessed at 3.2 years (range 2.4–5.1); 23 children were age 3 years or older. If the cerebellum was above the foramen magnum or had herniated to C2, C3, or C4, respectively, childhood seizures occurred in 0%, 7%, 21%, and 100%; high-risk bladder dysfunction in 33%, 33%, 71%, and 100%; and inability to ambulate independently in 20%, 70%, and 100% (no C4 cases) all P<.05. Ventriculomegaly, cerebellar herniation, and extraaxial space effacement were significantly associated with the need for ventriculoperitoneal shunt; however, 94% of children required shunt placement.
CONCLUSION: In fetuses with spina bifida, worsening cerebellar herniation on MRI was significantly associated with childhood seizure activity, high-risk bladder dysfunction, and lack of independent ambulation.
LEVEL OF EVIDENCE: II
Spina bifida, one of the most common disabling birth defects, affects approximately one in 1,000 pregnancies.1–3 It is associated with a range of central nervous system abnormalities, including complex hindbrain involvement known as the Chiari II malformation. The open neural tube is postulated to inhibit the accumulation of cerebrospinal fluid and pressure within the cranial vesicles, essential to normal brain development. As a result, a small posterior fossa, cerebral disorganization, and scalloping of the skull, or lückenschädel, occur.4 Children with spina bifida have varying degrees of paralysis, bladder and bowel morbidity, and developmental delay. To optimize the quality of life and functional outcome for these children requires lifelong coordinated care from experts in neurosurgery, urology, orthopedics, neurodevelopmental pediatrics, nutrition, and rehabilitation.5,6
Although the prenatal detection of spina bifida is excellent, the prognosis of affected children is highly variable, and efforts at predicting outcome have had limited success. Factors that have been associated with poor motor function and developmental delay include higher spinal lesion level and the presence of ventriculomegaly.7,8 However, because the majority of spina bifida lesions involve the lumbosacral region and do have accompanying ventriculomegaly, these factors may be limited as predictors. Fetal magnetic resonance imaging (MRI) has been increasingly used to improve visualization of relatively subtle anatomic abnormalities and has been particularly helpful in characterizing anomalies of the central nervous system.9 We hypothesized that findings encountered in the MRI evaluation of fetal spina bifida such as the degree of cerebellar herniation and effacement of the subarachnoid space might permit better understanding of the pathophysiology of fetal damage in these cases and potentially improve prediction of sequelae.
Our goal was to systematically review prenatal and postnatal findings of pregnancies diagnosed with fetal Chiari II malformation to characterize MRI central nervous system markers that may improve prediction of neonatal outcome.
MATERIALS AND METHODS
This was a review of pregnancies complicated by fetal spina bifida referred to a single institution for fetal MRI between January 1, 2001, and December 31, 2007. Pregnancies complicated by spina bifida have been routinely referred to our center for fetal MRI since 2001. All pregnant women whose children subsequently received interdisciplinary care at our regional spina bifida center were included. This center provides integrated care, free of charge, to all state residents up to 18 years of age with spina bifida coordinated through a multidisciplinary clinic. Pregnancies evaluated after 2007 were excluded, because it would have been difficult to assess meaningful outcomes in children younger than 1 year of age. Pregnancies complicated by cervical spine defects, other fetal anomalies or aneuploidy, and those participating in a fetal surgery trial were also excluded.
Pregnancies were ascertained by review of a fetal MRI database that is maintained for clinical operations and quality assurance. Gestational age was verified with sonography in all cases. Magnetic resonance images were reviewed by investigators blinded to outcome information (T.T.C. and D.M.T.). The atrium of the lateral ventricle was measured in the axial plane and was categorized as normal if less than 10 mm, mild ventriculomegaly if 10–15 mm, and severe ventriculomegaly if 15 mm or more. Extraaxial space effacement was also measured in millimeters in the axial plane above and below the level of the tentorium cerebelli. The extraaxial space above the tentorium was measured at the level of the precentral sulcus; the space below was measured at the level of the cisterna magnum. If there was no measurable space, it was considered to be completely effaced (Fig. 1A and B). The cerebellum was evaluated in the sagittal plane, and its degree of herniation was determined according to the lowest cervical vertebral level reached by the inferior aspect of the cerebellar vermis (Fig. 2A–D). The cerebellum was considered not to be herniated if the inferior border of the vermis was located above the foramen magnum. Subsequent levels were assigned as C2 (below the foramen magnum to the inferior margin of C2), C3 (below C2 to the inferior margin of C3), and C4 (below C3 to the inferior margin of C4).
Outcomes of infants and children were obtained from record review. The diagnosis of growth restriction was based on birth weight below the third percentile for gestational age using nomograms from our population that are adjusted for maternal ethnicity, infant sex, and nulliparity.10 Outcomes related specifically to spina bifida included ventriculoperitoneal shunt requirement, seizure activity, dysphagia, high-risk bladder dysfunction, and ambulation status. Dysphagia was diagnosed by formal swallow evaluation in children 1) whose parents had reported gagging or choking during feeding or 2) whose clinical histories suggested findings consistent with dysphagia, aspiration, or both. All children received formal urological evaluation, with urodynamic studies, cystometry, and renal ultrasonography in most cases. Children were diagnosed with high-risk bladder dysfunction based on the presence of renal damage or urodynamic test results that indicated increased risk for renal damage, including elevated bladder pressures, uninhibited bladder contractions, or both. For study purposes, the ability to ambulate independently, without a wheelchair or other devices, was assessed only in children 3 years of age or older.
Lesion level was assigned based on the highest vertebral level involved in a midsagittal image of the spine. Lesion level assignment was divided into one of three categories: T to L2, L3 to L4, or L5 to S1, as described by Johnson et al.11
Categorical data were reported as frequencies, and statistical significance was determined using χ2 analysis. Three or more group comparisons were analyzed using the Mantel-Haenszel test for linear trend.12 Logistic regression was used to adjust for spinal lesion level (highest level) in outcome analyses. P<.05 was considered statistically significant. Statistical analyses were performed using SAS 9 (SAS Institute, Cary, NC). The study was approved by the institutional review boards of the University of Texas Southwestern Medical Center and Texas Scottish Rite Hospital for Children.
Thirty-six mother–infant pairs were included. Demographic characteristics are presented in Table 1. Fetal MRI was performed at an average gestational age of 27 weeks (±6 weeks). The majority of neonates were delivered full-term with only three infants delivered preterm, each at 36 weeks. Two neonates had growth restriction. The mode of delivery for all but one neonate was by cesarean delivery. The median age of the children at the time of this study was 3.2 years (interquartile range 2.4–5.1 years). Lesion level is also presented in Table 1. Twenty-six of 36 fetuses (72%, 95% confidence interval [CI] 54–86%) had lumbosacral lesions with highest level L3 or below. Lesion level and vertebral segment span were significantly associated with each other in our patient cohort (Mantel-Haenszel test P=.002).
Overall, 34 children required ventriculoperitoneal shunt placement (94%, 95% CI 81–99%). Shunts were placed at the discretion of the attending pediatric neurosurgeon based on degree of ventriculomegaly after meningomyelocele repair. The shunts were placed within 3 weeks of birth for all but one child, whose shunt was placed at age 2. Five (14%, 95% CI 5–29%) developed seizures requiring treatment. Dysphagia was diagnosed by swallow study in 12 children (33%, 95% CI 19–51%). Urodynamic studies demonstrated neurogenic bladder in all 36 children and high-risk bladder dysfunction in 18 (50%, 95% CI 33–67%). Ambulatory status was assessed in the 23 children (64%) who were age 3 years or older. Seven children (30%) ambulated independently, seven (30%; 95% CI 13–53%) ambulated with devices, and nine (39%; 95% CI 20–61%) used a wheelchair.
In Table 2, outcomes are presented according to according to ventricular size at fetal MRI. The degree of ventriculomegaly—or more precisely, the presence of any ventriculomegaly—was significantly associated with the need for shunt placement (P=.006). Thirty-four neonates (94%) required a shunt, and neither of the neonates who did not require a shunt had evidence of ventriculomegaly at time of fetal MRI. The degree of ventriculomegaly was not significantly associated with any of the other adverse outcomes studied, and none of these relationships became statistically significant after adjusting for lesion level.
Table 3 demonstrates postnatal outcomes according to extraaxial space effacement on fetal MRI, divided into categories of 1) incomplete effacement above and below the tentorium, 2) incomplete effacement above and complete effacement below the tentorium, and 3) complete effacement above and below the tentorium. The degree of extraaxial space effacement was significantly associated with shunt requirement (P<.001). The two children without the need for a shunt had no evidence of complete effacement above or below the tentorium, ie, both fetuses had measurable pockets of extraaxial fluid in each plane. Extraaxial space effacement was not significantly associated with any of the other outcomes studied, and none of these relationships became statistically significant after adjusting for lesion level.
Outcomes according to the degree of cerebellar herniation seen on fetal MRI are presented in Table 4. With increasing severity of cerebellar herniation, there was a significant increase in shunt requirement, seizure activity, and high-risk bladder dysfunction. The two unshunted patients, who did not have ventriculomegaly or extraaxial space effacement, also had lesser degrees of cerebellar herniation (above the foramen magnum). Four of five children with seizures had a finding of cerebellar herniation below C3. High-risk bladder dysfunction was seen in 33%, 33%, 71%, and 100% of the children with herniation to above the foramen magnum, C2, C3, and C4, respectively. Children with lesser degrees of cerebellar herniation were more likely to ambulate without devices; in other words, those children with more advanced degrees of cerebellar herniation on prenatal MRI were more likely to require assistive equipment to ambulate independently. After adjusting for spinal lesion level using logistic regression, these outcome associations remained statistically significant.
The relationships among the three MRI variables were also evaluated because central nervous system abnormalities in fetuses with Chiari II malformation are likely to be part of a continuum. The degree of cerebellar herniation was significantly associated with increasing ventricular size, extraaxial space effacement above, and effacement below the tentorium cerebelli, all P<.05. This implies that the degree of cerebellar herniation was an independent predictor of seizure activity, high-risk bladder dysfunction, and lack of independent ambulation because no significant association was seen between ventriculomegaly and extraaxial space effacement and these adverse outcomes.
In our series, fetuses with spina bifida who had greater degrees of cerebellar herniation at MRI were at significantly greater risk for later development of seizure activity, high-risk bladder dysfunction, and lack of independent ambulation by the age of 3. These findings remained significant after adjusting for spinal lesion level. Cerebellar herniation, absence of ventriculomegaly, and incomplete effacement of the extraaxial space were also significantly associated with the need for a ventriculoperitoneal shunt; however, because the vast majority of children with spina bifida require shunt placement, prediction of this outcome is less likely to be helpful clinically.
It is known that the size of the cerebellum is reduced in children with Chiari II malformation, and this is believed the result of crowding of posterior fossa structures.13,14 Data are limited regarding the use of fetal MRI findings to predict outcomes in children with spina bifida, although the cerebellar herniation we report has been characterized using MRI, particularly in children. Using fetal MRI, Ando et al15 reported that the posterior fossa characteristically appeared “tight” in the setting of Chiari II malformation, although degree was not assessed or correlated with outcome. Salman et al13 performed MRI studies in 68 children with spina bifida and found significant herniation of the cerebellar vermis below the foramen magnum, a mean of 17 mm herniation compared with no herniation in control patients. Danzer et al16 described outcomes in 48 children who had undergone fetal myelomeningocele closure and reported that not only did MRI demonstrate significant improvement in hindbrain herniation in 47 children, but symptoms related to herniation were absent in more than 50%. Our series found that high-risk bladder dysfunction and ambulatory status were also associated with degree of cerebellar herniation.
In our series, 31 fetuses (86%) had evidence of ventriculomegaly, which is similar to other reports. Biggio et al17 found that 88% of 33 fetuses with spina bifida developed ventriculomegaly and that the majority did so by 21 weeks. Previous work has demonstrated that the threshold for ventriculomegaly diagnosed by MRI is the same as for ultrasonography and that ventricular size is independent of gestational age.18 However, Biggio reported that one third of patients did not develop ventriculomegaly until the third trimester. Therefore, it is possible that a larger series with serial assessment of ventricular size would be needed to fully characterize the relationship between development of ventriculomegaly and adverse outcomes in children with spina bifida. Biggio's group also evaluated ambulatory status according to ventriculomegaly and did not find a relationship.7 Peralta et al8 reported that ventriculomegaly was associated with poorer neurological development, something we could not address in our cohort.
Other limitations of this series are its relatively small size, its retrospective nature, and that it involved only fetal MRI. This was a feasibility pilot study because no data were available to estimate either the prevalence of the central nervous system findings we described nor their relationship to childhood outcomes. For this reason, no sample size was predetermined. Associations that were found to be not to be statistically significant may have become significant given a larger sample size. The sonographic detection of spina bifida is excellent,19 and it is our practice to characterize fetal spina bifida using both sonography and MRI. Because these patients were referred for our expertise in fetal MRI, we did not have the opportunity to evaluate them sonographically and thus we cannot address whether MRI modified the findings. Twickler et al9 reported that MRI can provide additional information in up to two thirds of fetuses with central nervous system abnormalities and that the optimal benefit of MRI may be its role as an adjunct to ultrasonography at later gestational ages. Saleem et al20 recently reported that fetal MRI findings changed the diagnosis in more than 40% of pregnancies referred for suspected fetal neural tube defects; however, the diagnosis was unchanged in all cases of myelomeningocele, and management-mode of delivery, was changed from vaginal delivery to cesarean delivery in one of nine patients.
Our finding that worsening degrees of cerebellar herniation were associated with multiple adverse outcomes in children with spina bifida may be helpful in counseling families and informing physicians caring for the child in postnatal follow-up and may indicate a role for fetal MRI in pregnancies complicated by spina bifida. Further study of whether the degree of cerebellar herniation can be reliably assessed sonographically in the setting of Chiari II malformation would be of interest. Future studies with MRI could include the relationship of central nervous system findings to other outcomes such as cognitive development.
1. Neural tube defects. ACOG Practice Bulletin No. 44. American College of Obstetricians and Gynecologists. Obstet Gynecol 2003;102:203–13.
2. Cunningham FG, Leveno KJ, Bloom SL, Hauth JC, Rouse DJ, Spong CY. Prenatal Diagnosis and Fetal Therapy. Williams Obstetrics. 23rd ed. New York (NY): McGraw-Hill, 2010. p. 287.
4. McLone DG, Knepper PA. The cause of Chiari II malformation: a unified theory. Pediatr Neurosci 1989;15:1–12.
5. Thompson DN. Postnatal management and outcome for neural tube defects including spina bifida and encephaloceles. Prenat Diagn 2009;29:412–9.
6. Bowman RM, Boshnjaku V, McLone DG. The changing incidence of myelomeningocele and its impact on pediatric neurosurgery: a review from the Children's Memorial Hospital. Childs Nerv Syst 2009;25:801–6.
7. Biggio JR Jr, Owen J, Wenstrom KD, Oakes WJ. Can prenatal ultrasound findings predict ambulatory status in fetuses with open spina bifida? Am J Obstet Gynecol 2001;185:1016–20.
8. Peralta CF, Bunduki V, Plese JP, Figueiredo EG, Miguelez J, Zugaib M. Association between prenatal sonographic findings and post-natal outcomes in 30 cases of isolated spina bifida aperta. Prenat Diagn 2003;23:311–4.
9. Twickler DM, Magee KP, Caire J, Zaretsky M, Fleckenstein JL, Ramus RM. Second-opinion magnetic resonance imaging for suspected fetal central nervous system abnormalities. Am J Obstet Gynecol 2003;188:492–6.
10. McIntire DD, Bloom SL, Casey BM, Leveno KJ. Birth weight in relation to morbidity and mortality among newborn infants. N Engl J Med 1999;340:1234–8.
11. Johnson MP, Gerdes M, Rintoul N, Pasquariello P, Melchionni J, Sutton LN, et al. Maternal–fetal surgery for myelomeningocele: neurodevelopmental outcomes at 2 years of age. Am J Obstet Gynecol 2006;194:1145–50; discussion 1150–2.
12. Mantel N, Haenszel W. Statistical aspects of the analysis of data from retrospective studies of disease. J Natl Cancer Inst 1959;22:719–48.
13. Salman MS, Blaser SE, Sharpe JA, Dennis M. Cerebellar vermis morphology in children with spina bifida and Chiari type II malformation. Childs Nerv Syst 2006;22:385–93.
14. Dennis M, Edelstein K, Hetherington R, Copeland K, Frederick J, Blaser SE, et al. Neurobiology of perceptual and motor timing in children with spina bifida in relation to cerebellar volume. Brain 2004;127:1292–301.
15. Ando K, Ishikura R, Ogawa M, Shakudo M, Tanaka H, Minagawa K, et al. MRI tight posterior fossa sign for prenatal diagnosis of Chiari type II malformation. Neuroradiology 2007;49:1033–9.
16. Danzer E, Finkel RS, Rintoul NE, Bebbington MW, Schwartz ES, Zarnow DM, et al. Reversal of hindbrain herniation after maternal-fetal surgery for myelomeningocele subsequently impacts brain stem function. Neuropediatrics 2008;39:359–62.
17. Biggio JR Jr, Wenstrom KD, Owen J. Fetal open spina bifida: a natural history of disease progression in utero. Prenat Diagn 2004;24:287–9.
18. Twickler DM, Reichel T, McIntire DD, Magee KP, Ramus RM. Fetal central nervous system ventricle and cisterna magna measurements by magnetic resonance imaging. Am J Obstet Gynecol 2002;187:927–31.
19. Dashe JS, Twickler DM, Santos-Ramos R, McIntire DD, Ramus RM. Alpha-fetoprotein detection of neural tube defects and the impact of standard ultrasound. Am J Obstet Gynecol 2006;195:1623–8.
20. Saleem SN, Said AH, Abdel-Raouf M, El-Kattan EA, Zaki MS, Madkour N, et al. Fetal MRI in the evaluation of fetuses referred for sonographically suspected neural tube defects (NTDs): impact on diagnosis and management decision. Neuroradiology 2009;51:761–72.
© 2010 The American College of Obstetricians and Gynecologists