Haemophilus influenzae and Streptococcus pneumoniae infections in children with cerebrospinal fluid shunts.

OBJECTIVE: This paper reviews the frequency of central nervous system infections due to Haemophilus influenzae and Streptococcus pneumoniae associated with cerebrospinal fluid (CSF) shunts in pediatric patients. The need for immunizations in this patient population is also evaluated. PATIENTS: All patients with cerebrospinal fluid shunts except those with brain tumors seen in our clinics. METHODS: We reviewed data in three computer databases, kept prospectively recording details of CSF shunt procedures and CSF shunt-related infections. RESULTS: 1,226 patients underwent 3,889 shunt placements between 1957 and 2007. Twelve patients had 14 episodes of Haemophilus or pneumococcal infections. CONCLUSIONS: Children with CSF shunts are at high risk for infection with H. influenzae and S. pneumoniae. Routine immunizations during infancy in addition to the 23-valent polysaccharide pneumococcal vaccine should be highly and actively encouraged by health care providers caring for children with CSF shunts. Additional expanded-coverage vaccines should be utilized if and when they become available.

Longevity of patients born with myelomeningocele.

There are limited data concerning the life expectancy for individuals born with myelomeningocele (MM), with and without hydrocephalus. To ascertain such data was our first purpose. We have selected all patients with MM in our computer database, The Patient Data Management System (PDMS/fx). Data were transferred to Excel for primary and SPSS/PC for final analysis by Kaplan-Meier life survival curves. Of the 1,054 patients with MM in the Birth Defects Clinic and the University of Washington Medical Center (UWMC) of Seattle, 505 are now over the age of 21 (391) or have died (114). Follow-up information was available since 1994 for 132, 62% of whom we have had contact within the past 2 years. The second purpose was to identify potential health factors associated with long-term outcome of patients with MM. Patient variables chosen as relevant to survival included hydrocephalus, treatment before or after 1975, and health maintenance determined by outcome for those receiving care within the last 5 years or those last seen before. Age at last appointment and reason for visit were determined in order to identify age-specific health care needs of the adult population. Survival and medical needs were obtained from the UWMC's computer database, Mindscape, and by telephone survey for adult patients not seen in the last 2 years. Death is more frequent earlier in life for those MM patients with hydrocephalus. Ordinary degenerative disorders affect MM patients earlier in life than normals. Our data extend life expectancy for patients with MM and hydrocephalus to age 40 years with some reliability for those treated from 1957 to 1974, but only 24 years for those treated with modern techniques after 1974. More data is needed to determine long-term survival.

19-year follow-up of fetal myelomeningocele brought to term.

We conducted a 19-year follow-up study of 223 infants with myelomeningocele born by cephalic presentation, 68 born by pre-rupture of amniotic membranes cesarean section (PRAM C/S) and 155 born vaginally to determine outcome. We recorded radiographic level as the last intact vertebral arch at the cephalad end of the spinal defect from standard AP films of the spine. Protrusions of the lesions dorsal to the back were obtained from medical records or parental recall. Subsequent spinal cord abnormalities (hydromyelia, tethered cord and hypoplasia) were determined by patient's symptoms and verified by MRI. Data were from our Patient Data Management System and analyzed using Microsoft Excel and Epi Info 6. Flat lesions protruding less than 1.0 cm beyond the plane of the back and those associated with breech position or kyphus were not benefited by PRAM CIS (Fisher's Exact, p = 0.58 between the two types of delivery for motor level). Flat lesions were more common in the vaginally delivered group (p = 0.01). Lesions protruding equal to 1.0 cm or beyond were associated with less paralysis after PRAM C/S (p = 0.01). Although tethered cord syndrome was more common in the PRAM C/S group (p = 0.02), there was no difference in the muscle strength loss between the PRAM C/S and the vaginally delivered groups. Symptomatic hydromyelia was equally distributed but spinal cord hypoplasia was more common in the PRAM C/S group (p = 0.03). Due to the limited numbers of patients in each group available for analysis, we recommend further study to ascertain the appropriate management of fetal myelomeningocele diagnosed in utero and brought to term.

Anatomy of the spinal cord in patients with meningomyelocele with and without hypoplasia or hydromyelia.

We compared the cross-sectional areas of spinal cords of patients with normal cords to the area of patients with meningomyelocele. The control group consisted of examinations of 27 patients with normal spinal cords providing 1547 axial images at 20 levels, C2-L2. The meningomyelocele group consisted of 67 MRI examinations of 41 patients, providing 4,095 axial images at 23 levels C2 to S1. Thirty-four examinations were of 21 patients with minimal hydromyelia, 7 examinations were of 3 patients with operable hydromyelia, and 26 examinations were of 17 patients without hydromyelia. In an additional analysis, we selected those meningomyelocele patients with cord tethering but without hydromyelia or hypoplasia (53 examinations of 30 patients) and compared them to symptomatic hypoplasia cases (9 examinations of 6 patients). The symptomatic hypoplasia cases were chosen because of progressive loss of muscle strength and worsening spasticity not relieved by surgical adhesiolysis. The test, retest error was 5.6% with differences between the means of repeated readings not being significant. All tests for significance were paired T test. The areas of spine levels C7-L2 for the controls were significantly larger than for the meningomyelocele patients (p = 0.000007). Including all levels C2-S1, the minimal hydromyelia cases were not significantly different from those without hydromyelia (p = 0.5). The areas C2-S1 of operable hydromyelia cases were larger than both non-shunted minimal hydromyelia (p = 0.00009) and of meningomyelocele patients without hydromyelia (p = 0.00003). The areas C7-L2, of hypoplasia cases were significantly smaller compared to the "normal" meningomyelocele cases (p = 0.0004). These data suggest that hydromyelia stimulates overgrowth of the cord, as does hydrocephalus of the brain, and that adhesiolysis procedures are of no value with hypoplasia of the spinal cord.

Vertebral column and spinal cord malformation in children with exstrophy of the cloaca, with emphasis on their functional correlates.

Exstrophy of the cloaca is a dramatic malformation whose embryology is poorly understood. While the management of this disorder has received significant attention in the urology and general surgery literature, the neurologic status of these children has been poorly addressed. In order to better characterize the spinal cord and vertebral column malformations found in children with exstrophy of the cloaca, we undertook a clinical review of 26 consecutive children with exstrophy of the cloaca who had been seen at a single institution over 28 years. The prevalence of vertebral malformations in the 25 children who could be evaluated was 25/25 (100%). Twenty (80%) of the children had at least one vertebral fusion, most frequently at T-7. Twenty-two (88%) of the children had at least one vertebra with deficient posterior elements, and the spinal levels most frequently involved were S-2, S-3, S-4 and S-5. Nine (36%) of the children had at least one vertebra with a narrowed interpedicular distance, most frequently at T-7. Nine (36%) of the children had at least one vertebra with atrophic facet anatomy, most frequently at L-3. The prevalence of myelodysplasia in the 19 children for whom spinal magnetic resonance imaging or intraoperative findings were available was 100%. Of these 19 children, 15 (79%) had myelocystocele, 2 (11%) had a lipomeningocele, 2 (11%) had a meningocele, 2 (11%) had hydromyelia, and 4 (21%) had a tethered cord. These data suggest that spinal cord and vertebral column malformations are very common in children with exstrophy of the cloaca.

Epidemiology of tethered cord with meningomyelocele.

This paper describes the epidemiology of tethered cord syndrome and its etiologies and co-morbidities following initial repair of both meningomyeloceles and lipomeningomyelocele. A review of the pertinent literature and data from 654 cases of meningomyelocele and 118 cases of lipomeningomyelocele has been drawn from a computerized database, Patient Data Management System/fx. Only cases born since 1964 were analyzed for the etiologies, co-morbidities, spinal cord abnormalities detected by contrast studies or MRI and for significant symptoms and signs. Tethered cord symptoms were related to an attachment to a rigid tether for all 31 cases following lipomeningomyelocele repair but 62 (75%) of the 83 post meningomyelocele repair patients developed the symptoms of tethered cord. Causes other than, or in addition to, tethering included an obstructed cerebrospinal fluid shunt, syringohydromyelia, benign tumor and spinal cord hypoplasia. Quantitative differentiation between asymptomatic thin spinal cords and symptomatic spinal cord hypoplasia as well as between central canal enlargement and symptomatic syringohydromyelia could not be demonstrated. Collaborative, multi-center studies of larger numbers of patients are recommended.

Plasma homocysteine and methionine concentrations in children with neural tube defects.

Mild to moderate homocysteinemia in women has been associated with an increased frequency of pregnancies with neural tube defects (NTD). Homocysteinemia is also an independent risk factor for premature vascular disease. In addition to folic acid, supplemental Vitamin B12, Vitamin B6 and betaine may normalize homocysteine metabolism, decrease the risk for NTD formation, and correct related metabolic imbalances in children with NTD. By means of automated amino acid analysis, we assessed total non-fasting homocysteine and methionine in plasma from 24 children with myelomeningocele. This study group (mean age 10.5 +/- 4.9 years) included 12 girls and 12 boys randomly selected from our Birth Defects Clinic. Homocysteine concentrations in our patients (4.7 +/- 1.8 mumol/L) did not differ from those of 20 randomly selected child controls (5.1 +/- 2.6 mumol/L). The mean homocysteine concentration for 36 adult controls (9.3 +/- 3.0 mumol/L) was significantly higher than the mean for either group of children (p < 0.0001). Linear regression analysis revealed negative correlation of total plasma homocysteine with serum folate (r = -0.53; p = 0.01), but not of homocysteine with either methionine or B12. Plasma methionine concentrations from our patients did not differ from adult reference values. Elevated homocysteine in some mothers of children with NTD has been attributed to defective methylation of homocysteine. These preliminary results do not indicate such a defect in the children themselves. A more comprehensive study of homocysteine, methionine and related metabolites in children with NTD and age-matched controls will be required to determine the clinical significance of these findings.

Comparison of erythrocyte antioxidant enzyme activities and embryologic level of neural tube defects.

Increased exposure to oxidant-derived free radicals or inadequate systems for antioxidant defense could alter cellular response at critical points in development. We measured 5 antioxidant enzymes, glutathione peroxidase (GSH-Px), glutathione reductase, glutathione-S-transferase, catalase and superoxide dismutase in erythrocytes and their plasma cofactor trace elements (Se, Zn, Cu) in 37 children with myelomeningocele and in 37 age-matched controls. We placed the patients into 3 groups according to motor level of the lesion at birth. We found significantly lower GSH-Px activities (p = 0.007) in children with myelomeningocele. For paired comparisons among the 3 patient groups and controls, there were significant differences (p < 0.05) between controls and both high (thoracic) and raid (lumbar) level embryologic lesions. The finding of antioxidant enzyme variations in our patients with myelomeningocele may indicate a role for abnormal oxidative metabolism in the development of this defect. The contribution of oxidative stress to human birth defects warrants investigation. We discuss potential relationships between oxidative stress and energy metabolism during primary neurulation.

Erythrocyte antioxidant enzyme activities in children with myelomeningocele.

Erythrocyte free radical scavenging enzyme activities and their cofactor trace elements in plasma were assessed in 26 selected patients with myelomeningocele, both parents from 10 selected families, and 14 healthy adult controls. All index children except one were deficient in erythrocyte glutathione peroxidase (GSH-Px). Nine of 10 parent pairs had at least one parent with deficient GSH-Px activity. Children with myelomeningocele had significantly lower GSH-Px activities than their parents; the group of 10 parent pairs had significantly lower GSH-Px activities than the control group; and glutathione reductase activities were significantly lower in parents and children with myelomeningocele compared with controls. A deficiency in one or more antioxidant enzymes may increase the risk for neural tube defects.

Epidemiology, etiologic factors, and prenatal diagnosis of open spinal dysraphism.

The caudal neural tube closes late in the first month after fertilization and failure of it results in myelomeningocele. Epidemiologic studies have shown differences in prevalence at birth based on ethnic-racial backgrounds and geography. Etiologic factors include the drug valproic acid or carbamazepine. Periconceptional folic acid supplementation appears to decrease the prevalence of neural tube defects. Numerous modalities allow for prenatal diagnosis of myelomeningocele. A cesarean section, before rupture of amniotic membranes and onset of labor, decreases the degree of paralysis.

Fetal aqueductal stenosis diagnosed sonographically: how grave is the prognosis?

OBJECTIVE: Aqueductal stenosis is a common cause of fetal hydrocephalus. Published studies of neonates with aqueductal stenosis have noted variable outcomes, with normal development seen in 24-86% of cases. In an attempt to better assess long-term outcomes in cases diagnosed in utero and to determine what prenatal sonographic findings might be used to predict prognosis, a retrospective analysis of patients with aqueductal stenosis was done. MATERIALS AND METHODS: Fifty-three consecutive cases of aqueductal stenosis discovered in utero at two high-risk obstetrical centers in Seattle between 1980 and 1993 were studied. Parents elected to continue pregnancy in 39 of these cases. Two months to 10 years of long-term follow-up was available in 30 patients, who form our study group. Prenatal sonograms, postnatal cranial ultrasound, and head CT and MR were evaluated. Prenatal sonographic data collected included the biparietal diameter, size of the lateral and third ventricles, the ratio of the two, and thickness of the frontoparietal cortical mantle. Medical records provided an assessment of development based on physical examination, meeting of major milestones, and neuropsychological testing. The in utero diagnosis of aqueductal stenosis was confirmed by postnatal CT, MR imaging, sonography, or autopsy. RESULTS: Within the study group of 30 patients, eight died in the postnatal period and four died subsequently. Of the 22 patients with adequate postnatal follow-up, moderate or severe developmental delay was present in 16 (73%). Normal developmental milestones were met in only three (10%) of all 30 patients and in 14% of those surviving the neonatal period. Although size of the lateral and third ventricles was not a useful predictor of long-term prognosis, the ratio of the two, as well as thickness of the frontoparietal cortical mantle, was weakly associated with long-term prognosis. No correlation was found between biparietal diameter and outcome. CONCLUSION: As compared with previous reports, prenatal diagnosis of aqueductal stenosis carries a grave prognosis. Twelve of the 30 patients died, for an overall mortality of 40%. Normal development was seen in only 10%, significantly less than in prior studies.

The outcome of fetal myelomeningocele brought to term.

14 years experience with 354 infants born with myelomeningocele are reviewed in view of prelabor and pre-rupture of amniotic membranes vs vaginal delivery or delivery by cesarian section after labor and rupture of amniotic membrane. The apparent lack of effect of severity of impairment determined by prenatal diagnosis on parental decision to carry their baby to term, the negative effect on motor function of kyphos or congenital kyphoscoliosis, the equivocal effect of breech presentation with or without rupture of amniotic membranes and the beneficial effect of prelabor and prerupture of membranes delivery of selected infants is discussed. The need for further multiple center, collaborative study to identify the contribution of a number of factors influencing the outcome of fetuses diagnosed as having myelomeningocele is emphasized.

Meningomyelocele: management in utero and post natum.

We report a four year follow-up of 39 of 47 infants born after pre-labour Caesarean section and 68 of 79 born vaginally. Loss of motor function due to late complications was more frequent in the Caesarean section group (Fisher's Exact; P = 0.004). However, the means of the differences between the X-ray levels (measured as the last intact vertebral arch seen on standard anteroposterior roentgenograms of the spinal column) subtracted from the motor levels still favour Caesarean section (mean = 3.24; SD = 2.7) over vaginal delivery (mean = 1.2; SD = 2.7) (Student's t-test; P = 0.0003). The frequencies of other complications, death and neonatal meningitis, were not significantly different. Another 38 infants born by Caesarean section after labour were more paralysed (mean of X-ray and motor difference = 1.8, SD = 2.2) following rupture of amniotic membranes than those with intact amniotic membranes with or without labour (mean = 3.4; SD = 2.2) (Student's t-test; P = 0.0067). The differences between X-ray and motor levels for patients born by Caesarean section with intact amniotic membranes and without labour (mean = 3.6; SD = 2.4) were not significantly different from those with labour and intact amniotic membranes (mean = 2.89; SD = 1.5). The number of new cases of meningomyelocele presenting to our clinic has decreased from an average of 30 per year between 1970 and 1987 to 14 between 1988 and 1992.

Cognitive abilities and lipomyelomeningocele.

Ten children with lipomyelomeningocele were evaluated with the WISC--R, the Wide Range Achievement Test--Revised, the Developmental Test of Visual-motor Integration, and the Child Behavior Checklist. These children were consecutive referrals to a birth defects clinic. Unlike their meningomyelocele counterparts, as a group these children appear to be average in their intellectual, academic, and behavioral characteristics. However, they exhibited low average perceptual motor skills, a feature more commonly seen in meningomyelocele.

The natural history of hip deformity in myelomeningocele.

We studied 1061 children with myelomeningocele, reviewing 3184 pelvic radiographs from 802 patients. Hip dislocation had occurred by the age of 11 years in 28% of children with a thoracic neurosegmental level, 30% of those with an L1/2 level, 36% of L3, 22% of L4, 7% of L5 and only 1% of those with sacral levels. Hip dislocation was not inevitable even when there was maximal muscle imbalance about the hip. The average hip flexion contracture in children aged 9 to 11 years was significantly greater in those with thoracic (22 degrees) and L1/2 (33 degrees) levels than in those with L4 (9 degrees), L5 (5 degrees) or sacral (4 degrees) levels. Our findings indicate that muscle imbalance is not a significant factor in the production of flexion deformity or dislocation of the hip; both are commonly seen in the absence of imbalance. The restoration of muscle balance should no longer be considered to be the principal aim of the management of the hip in children with myelomeningocele.

Mobility aids for children with high-level myelomeningocele: parapodium versus wheelchair.

The optimal method for providing mobility for children with myelomeningocele remains controversial. 39 children using a parapodium were compared with 29 children in wheelchairs. There were no significant differences between the two groups for medical complications, use of health-care services or activities of daily living, although there were differences in the patterns of complications. Children using the parapodium were more likely to develop lesions of the lower extremities, to have dislocated hips, to be more obese and to watch more television; children using wheelchairs were more likely to develop lesions of the gluteal region, to have knee-flexion contractures and to have fewer fecal accidents. The parapodium was judged by families to be less effective as a mobility aid; however, the upright posture it allows was considered extremely advantageous. A combined approach allowing upright posture and wheeled mobility would appear to be optimal.

Cognitive abilities and achievement status of children with myelomeningocele: a contemporary sample.

Designed study as a conceptual replication of Shaffer, Friedrich, Shurtleff, and Wolf (1985). Intelligence, school achievement, and perceptual motor skill data from 73 children with uncomplicated myelomeningocele were examined to determine their deviation from test norms. The respective impact of shunting and functional motor level was also assessed. The sample as a whole functioned below WISC-R and VMI norms. On the WRAT, deviation from the norms was evident only for the Arithmetic subtest. Partial replication of the Shunt X Motor level interaction reported by Shaffer et al. (1985) was obtained. Limitations of small-sample studies are discussed.

Modifications to the traditional description of neurosegmental innervation in myelomeningocele.

The consistency between patterns of lower-limb muscle strengths in children with myelomeningocele and Sharrard's classic description of segmental innervation was examined in 291 patients. Although it has been assumed that medial hamstring innervation was from a similar neurosegmental level as gluteus medius, and gluteus maximus was from a similar level as gastrocnemius-soleus, the authors found that medial hamstring strength more frequently correlated with iliopsoas and quadriceps, and glutei with anterior tibialis. It is proposed that children with myelomeningocele be grouped according to specific muscle strength rather than by neurosegmental level.