Materials and Methods

Retrospective chart and radiology reviews were performed for all infants who had undergone primary closure of a myelomeningocele defect at Children’s Hospital of Pittsburgh between the years of 1995 and 2015 after the University of Pittsburgh Institutional Review Board approval had been obtained (PRO11030590 and PRO8120394) on June 1, 2013. Eighty-three of these infants were born at Magee Women’s Hospital. Infants were matched with their birth mothers with the assistance of the Center for Assistance in Research with the use of eRecord of the University of Pittsburgh Medical Center. Infants who had been delivered outside of the University of Pittsburgh Medical Center system were excluded, as were those for whom records were incomplete and those who underwent in utero repair. Eighty infants had maternal data available. Three infants who were lost to follow up by the age of 2 years were excluded from the analysis. Records for 5 infants who died within 2 years of birth were analyzed separately from the remainder of the cohort, except where specified otherwise. All defects underwent standard neurosurgical closure within 24 hours of birth. The primary outcome measure was the motor function relative to the anatomic level of the myelomeningocele defect at the age of 2 years. The sample size that would be necessary to detect a difference between the groups with a power of 0.8 and a significance of .05 was calculated to be 52 subjects total (26 per group).

Pediatric data collection points appear in Table 1 . Maternal data points that were collected included acknowledged risk factors for fetal myelomeningocele (obesity, diabetes mellitus, previous pregnancy that was complicated by myelomeningocele, use of antiepileptic drugs, ethnicity), prenatal (preeclampsia, oligohydramnios, polyhydramnios, placental abruption), perinatal complications (hemorrhage, need for transfusion, surgical site infection or wound dehiscence, venous thromboembolism, pulmonary edema), and delivery method, as well as those listed in Supplemental Table 1 . Infant cohorts were grouped by delivery method, with those delivered via cesarean section subdivided into those delivered before and after the onset of labor. The groups were compared from the perspective of infant neurologic function relative to anatomic level of defect, with the use of the methods of Luthy et al. Labor was defined as regular contractions over a 2-hour period and cervical dilation of >2 cm. The anatomic level of the defect was defined as the most caudal intact lamina on x-ray, computed tomography, or magnetic resonance imaging, as previously defined. Luthy, Shurtleff, Liu, Cuppen, and Merrill Functional levels were labeled by the lowest level of purposeful movement on either side of the body at 24 months of age, as assessed on standard neurosurgical physical examination. This age was chosen because of its use in previous studies (Luthy, Shurlteff, Liu) and because motor function at age 2 years allows for the resolution of contractures that developed in utero and myelination of the corticospinal tracts that were not present in the immediate postnatal period, yet it is unusual for children of this age to have developed a symptomatic tethered cord with associated neurologic decline. The difference between anatomic and functional levels was analyzed by mode of delivery. The size of the myelomeningocele defect was defined as the largest diameter of the placode in centimeters measured at the time of closure. Independent sample t -tests were used to compare the groups for dependent variables that consisted of interval data. Pearson chi-square tests were used to compare groups for dependent variables that consisted of categoric data. Statistically significant differences were determined by probability values of <.05 (2-tailed). All statistical analyses were performed with the use of standard statistical software (IBM SPSS Statistics for Macintosh, Version 22.0., IBM Inc., Armonk, NY).

Results

Of the 72 infants for whom complete maternal and pediatric data were available, 22 infants were delivered vaginally, and 50 infants were delivered by cesarean section (44 before the onset of labor and 6 after the onset of labor; Table 1 ) The only maternal variable that reached statistical significance was an unsurprisingly longer hospitalization for the mother after cesarean delivery than after vaginal delivery ( Supplemental Table 1 ). Infant variables that reached statistical significance included a smaller myelomeningocele defect and a longer length of stay for the infant delivered vaginally. Infants who were delivered vaginally were more likely to have a seizure disorder ( Table 1 ).

The 5 infant deaths are included in the assessment of pregnancy risk factors and complications, for a total of 77 mothers. Sixty-four of the 77 women (83%) were white, 8 women (10%) were African American, and 1 woman (1%) was Asian. There were no Hispanic mothers, although there were 2 Hispanic fathers. The main maternal risk factor for myelomeningocele in this series was an elevated body mass index. Forty-five women (58%) were obese (body mass index, >30 kg/m ² ), and 19 women (25%) were overweight (body mass index, 25-29.9 kg/m ² ). Six women (8%) had gestational diabetes mellitus, and 6 women (8%) were taking antiepileptic drugs. Two women (3%) had pregestational diabetes mellitus, and 1 woman (1%) had a family history of myelomeningocele. Some mothers had >1 risk factor. The incidence of maternal complication of cesarean delivery was 0%. There were no maternal deaths.

Thirty-five of the 77 mothers (45%) had a complication of pregnancy other than the fetal myelomeningocele and its associated conditions, and 10 mothers (13%) had multiple complications. Complications included gestational diabetes mellitus (6 mothers), seizure disorder (6 mothers), gestational hypertension (5 mothers), preeclampsia (5 mothers), oligohydramnios (4 mothers), triplet gestation (3 mothers), polysubstance abuse (3 mothers), bipolar disorder (3 mothers), polyhydramnios (2 mothers), intrauterine growth restriction (2 mothers), nonreassuring fetal status (2 mothers), preterm delivery (16 mothers), HELLP (hemolysis, elevated liver enzymes, low platelets) syndrome (1 mother), thrombocytopenia (1 mother), and twin gestation (1 mother).

Fifty-five infants underwent cesarean delivery, which included the 5 infant deaths. Indications included elective (21 infants), repeat (13 infants), breech presentation (8 infants), multiple gestation (4 infants), massive hydrocephalus (3 infants), preeclampsia (2 infants), intrauterine growth restriction (2 infants), oligohydramnios (2 infants), nonreassuring fetal status (2 infants), and active herpes (1 infant). Some infants had >1 indication for cesarean delivery.

Of the 77 pediatric patients (including the 5 infant deaths), 11 patients (14%) underwent Chiari II decompression. Three of these Chiari II decompressions occurred in infants that died before the age of 1 year (60% of these patients), and the remaining 8 decompressions (11% of these patients) occurred in long-term survivors. Of the 72 long-term survivors, 10 pediatric patients (14%) underwent tethered cord release ( Table 1 ). Only 7 children (10%) did not undergo shunt placement, for a shunt placement rate of 90%. These rates are similar to previously reported rates, and route of delivery had no effect on them ( Table 1 ).

Shunt placement rates ( Table 2 ) were analyzed by mode of delivery and anatomic level. The overall rate of shunt placement was 90%. There was a statistically significant trend toward higher rates of shunt placement with lower levels of myelomeningocele defects, contrary to previous published reports. There was no relationship between rate of shunt placement and mode of delivery or presence of labor. The level of the defect correlated with the mode of delivery because of selection bias. Mothers who carried fetuses with lower defect levels were allowed to labor and deliver vaginally more often than those with higher defect levels, which is a confounding variable that becomes apparent on significance testing (chi-square test [5; n=72]=13.01; P =.02; Table 2 ). Notably, however, there was no correlation between mode of delivery and a difference between anatomic and functional outcomes.

Five pediatric patients (6.9%) died before 1 year of age ( Table 3 ). All of these infants were delivered by cesarean section without a trial of labor. The deaths are excluded from the calculations for Table 1 . All infants who died had central respiratory abnormalities (stridor from vocal cord paralysis [1 infant] or central apnea [4 infants]) that were apparent at birth, and severe hydrocephalus (with an average occipitofrontal circumference nearly 10 cm larger than that of the long-term survivors) warranted shunt placement on the day of myelomeningocele closure. None of the patients who died reached 37 weeks of gestation before delivery; the average age at delivery was 35.6 weeks, nearly 2 weeks earlier than the average age for infants who survived long term. Three of the 5 infants (60%) underwent Chiari II decompression shortly after birth, and the other 2 infants were too critically ill to allow for this. Four infants died after a parental decision to withdraw care; the fifth experienced Enterobacter sepsis and abdominal compartment syndrome and died after an unsuccessful attempt at resuscitation on day 9 of life. One infant was quadraparetic at birth, which is a recognized symptom of Chiari II malformation and thought to be due to compression or malformation of the corticospinal tracts that innervate the upper extremities or to the presence of a high cervical syrinx (which this patient did not have). Chiari II decompression had no respiratory or swallowing dysfunction effect on any of these infants.

All 8 fetuses in breech presentation were delivered by cesarean section in the absence of labor. One of the 8 died in the perinatal period; this patient’s data figure only into the immediate postnatal calculations and are excluded from the ambulation, anatomic-functional, Chiari II, and tethered cord assessments. The breech infants were more likely to have lower Apgar scores at 1 minute, larger head circumference that required shunting (and shunting at younger ages), cardiac and renal defects (although interestingly, not orthopedic defects), and nonambulatory status and to have undergone a tethered cord release, although only the larger head circumference reached statistical significance ( Table 4 ).

Functional levels were slightly lower than anatomic levels or slightly better than predicted for all groups. There was a trend toward even better function than predicted by anatomic level in the cohorts of infants who were born vaginally and after a trial of labor, contrary to a number of previous studies. There was no effect of mode of delivery or presence of labor on ambulatory status ( Table 5 ). A regression analysis was performed that examined the relationship between anatomic level, occipitofrontal circumference, size of defect, gestational age, birthweight, and maternal age to our outcome variables and mode of delivery and anatomic-functional difference at age 2 years, because these variables were thought to be most likely to have an impact on the outcomes that were measured. This analysis indicated that size of defect significantly predicted both of our outcome variables. Specifically, larger defect size was more likely to be associated with a cesarean delivery (β=.27; t[62]=2.16; P =.04); defect size also explained a significant proportion of variance in mode of delivery ( R ² =0.07; F[1,62]=4.67; P =.04). In addition, larger defect size was associated with a smaller anatomic-functional difference at age 2 years (β=–.29; t[62]=–2.39; P =.02); defect size also explained a significant proportion of variance in anatomic-functional difference ( R ² =0.09; F[1,62]=5.73; P =.02).

Table 5

Effect of mode of delivery and labor on neurologic function and ambulatory status

Variable	Delivery mode		P value	Presence of labor		P value
	Vaginal (n=22)	Cesarean (n=50)		Labor (n=28)	No labor (n=44)
Anatomic level (cervical, thoracic, and L1=0, L2-5=2-5, respectively; S1 and S2=6) a	4.4±1.7	3.7±1.6	.10	4.3±1.7	3.6±1.6	.13
Functional level (no motor function=0, iliopsoas=2, quadriceps=3, hamstrings=4, dorsiflexion=5, and plantar flexion=6) a	4.6±1.6	4.0±1.7	.22	4.4±1.6	4.0±1.7	.31
Difference between atomic and functional levels	–0.18±1.9	–0.48±1.5	.47	–0.18±1.8	–0.52±1.5	.38
Ambulate independently, n (%)	6 (27)	12 (24)	.77	8 (29)	10 (23)	.58
Ambulate with assistive device (orthoses, crutches, or walker), n (%)	11 (50)	28 (56)	.64	13 (46)	25 (57)	.57
Not ambulatory, n (%)	4 (18)	10 (20)	.95	7 (25)	10 (23)	.89

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related

Related posts:

1. Antenatal corticosteroids beyond 34 weeks gestation: What do we do now?
2. Urinary congophilia in women with hypertensive disorders of pregnancy and preexisting proteinuria or hypertension
3. The effect of combined inositol supplementation on maternal metabolic profile in pregnancies complicated by metabolic syndrome and obesity
4. Lithopedion: an unexpected finding

Stay updated, free articles. Join our Telegram channel

Join