Myelomeningocele Surgery over the 10 Years Following the MOMS Trial: A Systematic Review of Outcomes in Prenatal versus Postnatal Surgical Repair

Background and Objectives: Myelomeningocele is the most severe form of spina bifida, a congenital neural tube defect arising from an incomplete neural tube closure during early development with damage worsening with advancing gestational age. The Management of Myelomeningocele Study (MOMS) Trial proved that surgery performed before 26 weeks of gestation significantly improved the prognosis, significantly changing treatment paradigms. This article aims to provide a review of the changes and updates in spina bifida repair over the 10-year period following the MOMS Trial. Material and methods: We performed a systematic review in the PubMed and Cochrane databases as well as a hand-search of high-impact journals using the reference list of all identified articles, searching for randomized controlled trials and observational studies. Results: We identified 27 articles published between 2011 and 2021 that fulfilled the inclusion criteria and review them in the present study. Conclusions: With growing experience and with the improvement of prenatal open and fetoscopic techniques, the outcome of SB-associated conditions could be improved and the risks to both the mother and the fetus reduced. A continuous follow-up of the treated infants and further randomized trials are essential to study the complications and advantages or disadvantages of any given treatment strategy.


Introduction
Myelomeningocele (MMC), the most severe form of spina bifida (SB), is a congenital neural tube defect occurring with an incidence of approximately 1 in 2900 live births, arising during early development from an incomplete neural tube closure leading to an open spinal canal. The exposed spinal cord undergoes significant damage early in pregnancy due to the intrauterine environment with the spinal injury extending cranially and worsening with advancing gestational age [1,2], followed by a suction gradient on the hindbrain [3]. These events account for the basis of the "two-hit" hypothesis, the rationale behind prenatal surgery for an early MMC closure. The Management of Myelomeningocele Study (MOMS) Trial published in 2011 proved that surgery performed before 26 weeks of gestation significantly improved the prognosis by decreasing the risk of death or need for shunting by 12 months of age, reducing the degree of hindbrain herniation associated with Chiari II malformation (CM-II) and improving motor function and the likelihood of independent walking compared with postnatal surgery [4], thus significantly changing the treatment paradigms. This article aims to provide a review of the changes and updates in spina bifida repair over the 10-year period following the MOMS trial and to discuss the evidence of the superiority of prenatal versus postnatal surgical treatment in regard to both short-term and long-term prognoses.

Materials and Methods
We performed a systematic review in the PubMed and Cochrane databases as well as a hand-search of high-impact journals using the reference list of all identified articles, searching for randomized controlled trials and observational studies. The terms used for the search were "spinal dysraphism", "spina bifida" or "myelomeningocele" combined with "surgery" or "f(o)etoscopy". To be eligible for inclusion, studies had to be published between 2011 and 2021, written in English and reporting singleton fetuses with isolated spina bifida who underwent either a prenatal fetoscopic or open repair or postnatal surgical closure with a minimum of 30 days follow-up. Exclusion criteria were unrelated, duplicated, unavailable full-text or abstract-only papers. The primary outcome was perinatal mortality i.e., deaths during the fetal and postnatal (within the first month of life) periods. The secondary outcomes were maternal, fetal, neonatal and infant outcomes. The articles were categorized as either early experience or later experience, using a cutoff of 30 cases as proposed by Kohl et al. [5]. Nonrandomized studies were assessed for the risk of bias using the Newcastle-Ottawa Scale and judged on three perspectives: the selection of the study groups, the comparability of the groups and the outcome of interest [6]. The statistical significance was determined using the chi-squared and t-test statistics; p values < 0.05 were considered statistically significant.

Description of Studies
Our preliminary literature search identified 1869 publications; we used the PRISMA algorithm to assess them [7]. Duplicates were excluded, leaving us with 921 articles out of which 809 were excluded on reading the title and abstract. The remaining 112 were assessed for eligibility by full-text reading. The selection left us with 27 eligible articles that were included in the study ( Figure 1) .
decreasing the risk of death or need for shunting by 12 months of age, reducing the degree of hindbrain herniation associated with Chiari II malformation (CM-II) and improving motor function and the likelihood of independent walking compared with postnatal surgery [4], thus significantly changing the treatment paradigms. This article aims to provide a review of the changes and updates in spina bifida repair over the 10-year period following the MOMS trial and to discuss the evidence of the superiority of prenatal versus postnatal surgical treatment in regard to both short-term and long-term prognoses.

Materials and Methods
We performed a systematic review in the PubMed and Cochrane databases as well as a hand-search of high-impact journals using the reference list of all identified articles, searching for randomized controlled trials and observational studies. The terms used for the search were "spinal dysraphism", "spina bifida" or "myelomeningocele" combined with "surgery" or "f(o)etoscopy". To be eligible for inclusion, studies had to be published between 2011 and 2021, written in English and reporting singleton fetuses with isolated spina bifida who underwent either a prenatal fetoscopic or open repair or postnatal surgical closure with a minimum of 30 days follow-up. Exclusion criteria were unrelated, duplicated, unavailable full-text or abstract-only papers. The primary outcome was perinatal mortality i.e., deaths during the fetal and postnatal (within the first month of life) periods. The secondary outcomes were maternal, fetal, neonatal and infant outcomes. The articles were categorized as either early experience or later experience, using a cutoff of 30 cases as proposed by Kohl et al. [5]. Nonrandomized studies were assessed for the risk of bias using the Newcastle-Ottawa Scale and judged on three perspectives: the selection of the study groups, the comparability of the groups and the outcome of interest [6]. The statistical significance was determined using the chi-squared and t-test statistics; p values < 0.05 were considered statistically significant.

Description of Studies
Our preliminary literature search identified 1869 publications; we used the PRISMA algorithm to assess them [7]. Duplicates were excluded, leaving us with 921 articles out of which 809 were excluded on reading the title and abstract. The remaining 112 were assessed for eligibility by full-text reading. The selection left us with 27 eligible articles that were included in the study ( Figure 1) .   The articles were categorized as either early experience or later experience, using a cutoff of 30 cases as proposed by Kohl et al. [5] (Table 1). In the Table, we have listed the articles in order based on the year of publication starting from the oldest to the latest.

Mortality Rate
The comparison between infant mortality rates in different neurosurgery centers is summarized in Table 6. The mortality rate in prenatal series was not high (between 2% and 20%). Many postnatal series reported no immediate infant mortality.

Comparison between Surgical Techniques
The comparison between surgical techniques in different centers is summarized in Table 7. Out of the 25 articles that reported a single surgical technique, 12 (48%) reported a postnatal closure of the defect, 6 (27.3%) reported open prenatal surgery, 5 (20%) reported fetoscopic prenatal surgery and 2 (9%) reported a prenatal surgical intervention not otherwise specified. Pastuszka et al. [27] and Flanders et al. [13] reported both open prenatal and postnatal surgical interventions.

Comparison between the Results of the Reported Approaches
The studies included reported on many different outcomes. For a few of these outcomes, a comparison between the results obtained in specific management scenarios (prenatal versus postnatal treatment, early experience versus late experience) is possible. Overall, mortality was higher in prenatally-treated than in postnatally-treated cases (p < 0.01). Hydrocephalus and Chiari malformations were more frequent in the postnatally-treated than in the prenatally-treated cases (p < 0.05 for both outcomes). The rupture of membranes was less frequent and the gestational age at delivery was higher in late experience versus early experience series (p < 0.01 for both). Many outcomes were reported only by a few studies. The follow-up interval was different in different studies.

Discussion
A large amount of heterogenous data has lately been published on the in utero repair of spina bifida. As these data on fetal surgery accumulate, the results are judged against those of the conventional postnatal surgery for spina bifida. This analysis does not seem to be close to any conclusions and the approach that would benefit patients most is still not clearly known. Systematic reviews and meta-analyses are useful tools for drawing practical conclusions and making sense of heterogenous literature.
The history of surgical interventions for myelomeningocele starts with a series of trials and errors from a time when the rules regarding who and when should operate were not clearly defined. Many strategies have been tried over the years such as ligatures, injections, serial tapping and excision. While these early trials proved ineffective, they served to build the knowledge of the pathology involved and eventually resulted in modern approaches [35].
Miled et al. studied the timewise progression and topographic progression of neuronal loss in 186 cases of myelomeningocele and reported that a significant neuronal loss is present earlier than a gestational age of 16 weeks and progressively extends cranially, thus suggesting that an earlier prenatal repair could prevent Chiari II malformation, rescue the remaining motor neurons in the exposed cord and prevent the extension to the upper levels of the spinal cord [2]. Diagnosing fetal myelomeningocele can be done during routine antenatal appointments. Munoz et al. compared a prenatal ultrasound evaluation to a prenatal MRI and stated that MRI is not superior to ultrasound in the diagnosis of open spina bifida [36]. Recent advances in ultrasound technology allow for an early diagnosis at 12-14 weeks' gestation of severe spina bifida [37]. Spina bifida is sometimes associated with genetic anomalies such as trisomy 18. Both major structural defects and common aneuploidies can be routinely diagnosed in the first trimester of pregnancy [38,39].
The MOMS Trial, published in 2011 in the New England Journal of Medicine, was the largest randomized trial on the prenatal versus the postnatal treatment of myelomeningocele and aimed to confirm the benefits of in utero surgical interventions earlier than 26 gestational weeks and to assess the risks of such interventions [4]. According to the MOMS Trial, prenatal surgery for myelomeningocele significantly reduced the need for a cerebrospinal fluid shunt and significantly improved motor function at 30 months but came with costs in terms of both maternal and fetal morbidity. The inclusion criteria were a maternal age of over 18 years, US residency, a singleton pregnancy, an upper myelomeningocele boundary between T1 and S1, a hindbrain herniation, a gestational age between 19.0 and 25.9 weeks and a normal karyotype. The exclusion criteria were either regarding the mother such as a contraindication to surgery (e.g., a previous hysterotomy), a body mass index over 35 and a risk of preterm birth and placental abruption or regarding the fetus such as severe kyphosis or a fetal anomaly unrelated to myelomeningocele [4].
Antiel et al. studied the impact on family and parental stress of prenatal and postnatal repairs of myelomeningocele and reported that the overall negative parental impact of caring for a child with myelomeningocele was significantly lower in the antenatal treatment group compared with the postnatal treatment group, with the ambulation status and family resources being predictive of the impact on parental and family stress [40].
The results of the MOMS trial are clearly encouraging [4,41] but historically they have to be considered as early experience with fetal surgery. Research on the in utero repair of spina bifida continues and several fetal medicine centers have published their 'post-MOMS' experience on fetal surgery for spina bifida . More importantly, the postnatal repair of spina bifida carries on, for several obvious reasons, in many pediatric neurosurgery centers and yields good results [42]. In our study, 50% of the articles still reported a postnatal closure of the defect in cases of myelomeningocele. One of the latest studies included in our analysis [32] challenged the idea that the results of a prenatal repair of SB are significantly better than those of a postnatal repair. The study showed that the tethering of the spinal cord occurs at a higher rate in prenatally-treated patients than in postnatally-treated patients [32].
In the post-MOMS era, the therapeutic landscape has become more complex with the emergence of new (fetoscopic) techniques and approaches for in utero surgery for spina bifida.
Joyeux et al. in 2015 [43] and Kabagambe et al. in 2018 [44] compared a fetoscopic and an open repair of myelomeningocele and stated that the newer fetoscopic method takes longer to complete, has a greater risk of prematurity and requires additional postnatal procedures but has a comparable shunt rate and is not linked to uterine thinning or dehiscence [43,44]. However, larger studies and long-term data are needed.
Reviews of the literature on spina bifida management have been published over the entire decade since the MOMS trial [45][46][47]. A consensus statement of the Congress of Neurological Surgeons was based on the literature up to 2016 and was very much influenced by the initial results of the MOMS trial [45]. The methodology of systematic reviews varies widely and is sometimes overly stringent. For example, a meta-analysis from 2019 included only two studies out of all available literature: the 30-month results of the MOMS trial and a prospective cohort published in 2014 [47]. In this context, we consider our review not superfluous.
The 27 studies included in our review heterogeneously report on many different outcomes. Unfortunately, many outcomes were reported only by a few studies with sometimes conflicting results. For instance, of the two studies containing data on the gestational age at delivery in fetoscopic prenatal surgery, one reported an extremely low (32 weeks) mean gestational age at delivery [28] whereas the other reported an extremely high (38 weeks) mean gestational age at delivery [19]. The follow-up interval varied widely among the studies; therefore, a meaningful analysis of the results such as the ability to walk or retethering is difficult. We feel that a substantial statistical analysis of the prognosis associated with different treatment strategies is not possible because of the lack of uniformity of data reporting. Bias can be introduced by the retrospective design of the majority of studies. Interpretations of the data should be cautious.

Conclusions
Although there have been 10 years since the MOMS Trial, experience still needs to be gained and prenatal techniques need to be improved in order to obtain a better prognosis and lower risks for both the mother and fetus. Despite the apparent benefits of fetal surgery, only a few centers offer this technique, which is more technically demanding and requires a multidisciplinary effort [48]. With growing experience and with the improvement of prenatal open and fetoscopic techniques, the outcome of SB-associated conditions could be improved and the risks to both the mother and the fetus reduced. A continuous follow-up of the treated infants and further randomized trials with a uniform design are essential to study the complications and advantages or disadvantages of any given treatment strategy.