Neurosurgeons’ opinions on the prenatal management of myelomeningocele

OBJECTIVE Improvements in imaging and surgical technological innovations have led to the increasing implementation of fetal surgical techniques. Open fetal surgery has demonstrated more favorable clinical outcomes in children born with open myelomeningocele (MMC) than those following postnatal repair. However, primarily because of maternal risks but also because of fetal risks, fetal surgery for MMC remains controversial. Here, the authors evaluated the contemporary management of MMC in the hope of identifying barriers and facilitators for neurosurgeons in providing fetal surgery for MMC

M yeloMeningocele (MMC), characterized by ex- trusion of the spinal cord together with the meningeal membranes, is the most severe and common form of spina bifida. 11Standard treatment is closure of the MMC defect within 48 hours of birth to minimize infectious and traumatic risk to the tissue. 14,16Since publication of the first case of surgical closure of an MMC defect in 1892, only limited progress has been made in surgical repair with the primary goals of creating a watertight closure over the open neural placode. 6Patients born with MMC can face lifelong disabilities such as neurological deficits below the level of the defect, hydrocephalus, hindbrain herniation, scoliosis, tethering of the spinal cord, urinary and fecal incontinence, and sexual dysfunction.
Early in vivo studies suggested that prenatal intervention could prevent secondary insults to the neural placode and the development of hindbrain herniation or hydrocephalus due to persistent intrauterine CSF leakage.Further surgical innovations and improvements in instrumentation refined this technique and improved mortality and morbidity rates. 1,25Initially, fetal surgery served as an option to treat potentially lethal conditions such as congenital diaphragmatic hernia, improving early mortality rates. 19,28With increasing knowledge and experience, the indications for fetal surgery have broadened toward treating nonlethal diseases such as MMC. 1,8At the end of the last century, early case studies of fetal MMC repair showed promising results and eventually led to a randomized controlled trial published in 2011.
In the Management of Myelomeningocele Study (MOMS), women diagnosed with a fetus having MMC were randomized to undergo either open fetal surgery before 26 weeks of gestation or conventional postnatal surgical repair. 2 This trial was stopped prematurely given the efficacy of prenatal surgery.Results showed that only 40% of the prenatal surgery group needed placement of a shunt versus 82% in the postnatal surgery group at 1 year.In addition, children with MMC in the prenatal surgery group were less likely to have hindbrain herniation and were more likely to have a better level of motor functioning and to walk independently.However, these promising fetal outcomes came at the expense of maternal morbidity with higher rates of preterm delivery and a higher risk of maternal complications such as placental abruption and chorioamniotic membrane separation.Fetal surgery also carries risks for the fetus such as the need for transfusions and a lower birth weight.Furthermore, the MOMS had strict inclusion criteria, reflected by the fact that of the 1087 women who had undergone preliminary screening, only 183 underwent randomization.
Fetoscopic techniques were developed in the hopes of lowering maternal morbidity both in the current pregnancy and in future pregnancies.Developments and refinements have further improved fetoscopic techniques. 9,22nlike open prenatal repair, fetoscopic surgery was less well received by the neurosurgical community because of the complexity of the procedure and less-established outcomes. 12,13ith the rise of fetal surgery and the increasing evidence of its risks and benefits in treating selected pregnancies with MMC, neurosurgeons may increasingly face ethical dilemmas, such as weighing maternal risks against potential fetal benefits.An expected paradigm shift in the treatment of MMC since publication of the MOMS results has led to an interest in knowing about the current management and application of fetal surgery by our colleagues.We aimed to identify the barriers and facilitators of fetal closure perceived by the neurosurgeons providing current techniques for MMC repair.Therefore, we conducted a survey on the management of MMC by neurosurgeons to evaluate the clinical practice worldwide.

Methods
8,27 This survey was reviewed by two pediatric neurosurgeons experienced in treating MMC as well as a maternal-fetal medicine specialist before being finalized.The survey consists of 17 questions and can be divided into three sections: 1) characteristics of the respondents, such as subspecialization, tenure, and country of employment; 2) the practice of counseling on and managing MMC, such as the items discussed during counseling of parents, number of MMC defects closed yearly, and the availability of prenatal surgery to treat MMC; and 3) barriers, facilitators, and expectations of fetal surgery for MMC.The survey can be found in the Supplemental Material: Survey.The institutional review board waived the need for study approval as this research does not involve patients.
Members of the Congress of Neurological Surgeons (CNS) and the International Society for Pediatric Neurosurgery (ISPN) were approached by an email invitation containing a cover letter asking them to fill out a survey.The CNS was founded in 1951 and represents more than 9000 international members, who include residents, neurosurgeons, and medical students, among others.The ISPN consists of more than 350 pediatric neurosurgeons with the aim of improving the neurosurgical care of children.
In March 2019 the survey was distributed by email, using SurveyMonkey.To increase the response rate, reminders were sent.By filling out a survey, respondents gave informed consent.Only responses from neurosurgeons and neurosurgery residents were included.
Data were analyzed using the IBM SPSS Statistics, version 21.0 for Windows (IBM Corp.).Descriptive statistics were used to present data in frequencies and percentages.Categorical data were analyzed using the chi-square test.A p value < 0.05 was considered to be statistically significant.

Characteristics of the Respondents
A total of 448 respondents filled out the survey.Two responses were not from neurosurgeons or neurosurgery residents and were therefore excluded.Eighty-four percent of the respondents were from the CNS, while the remaining 16% were from the ISPN.Respondents were employed in a total of 58 countries, with the US (56.3%),Brazil (5.2%), and India (4.9%) having the most respondents (Fig. 1).Africa was the least represented (2.7%; Table 1).
Respondents had a mean clinical experience of 21.6 ± 12.9 years (mean ± standard deviation), with residents having a mean clinical experience of 5.5 years (Table 1).Most of the respondents had pediatric neurosurgery as their subspecialization (59.2%).Spine (29.6%) and neurooncology (28.3%) followed as the most frequent subspecializations.
Among the respondents, 81.0% treated patients with MMC.Neurosurgeons subspecialized in pediatric neurosurgery were more likely to treat patients with MMC (p < 0.001).Among neurosurgeons not subspecialized in pediatric neurosurgery, 54.8% treated patients with MMC.Neurosurgeons performing closure of MMC defects operated on a mean of 9.6 ± 17.9 patients annually.

Counseling and Management of MMC
Three-quarters (74.5%) of the neurosurgeons and 50% of the residents deemed themselves sufficiently informed to counsel parents (p = 0.008), and 72.2% of the respondents actually counsel potential parents with children suspected to have MMC.Of those neurosurgeons who deemed themselves insufficiently informed to counsel, 37.2% still counsel potential parents.Table 2 gives an overview of several items discussed by respondents during the counseling of potential parents.All items differed statistically significantly between pediatric and nonpediatric neurosurgeons.Surgical closure of the MMC defect, the need for CSF shunting, and the expected level of motor function were always discussed by 83.1%, 75.3% and 69.4% of respondents, respectively.Termination of the pregnancy and the option of fetal surgery were never discussed by 35.6% and 23.7% of the neurosurgeons, respectively.Ninety-one percent of the departments in which the neurosurgeons are employed provide conventional postnatal surgical closure of the MMC defect.Open fetal and fetoscopic procedures are provided by 13.0% and 4.9% of the centers, respectively.Four out of all centers provided fetal surgery but did not perform conventional postnatal surgery.Performing more surgical closures annually was not associated with providing fetal surgery (p > 0.05).
Table 3 gives an overview of the disciplines involved in the treatment of (unborn) patients with MMC.As regards clinical management by the respondents, neurosurgery (78.0%), obstetrics (77.4%), and pediatrics (59.9%) were always involved in the care of MMC patients.Rehabilitation medicine, orthopedics, and neurology were never involved according to 27.5%, 26.6%, and 23.4% of respondents, respectively.Pediatric neurosurgeons were less likely than nonpediatric neurosurgeons (p = 0.002) to involve pediatrics.

Barriers, Facilitators, and Expectations of Fetal Surgery for MMC
When asked to rank the most important objections to providing open fetal surgery to treat MMC, 33.8% ranked "too few cases available to become proficient in the technique" as number one (Fig. 2).The risk of maternal complications and the risk of fetal complications were ranked second and third, respectively, by 23.6% and 15.2%.Costs and ethical concerns were ranked last by 32.4% and 27.0%, respectively.
The most important reason to provide open fetal surgery according to respondents was a decreased rate of shunt dependency (37.8% of respondents; Fig. 3), followed by a decreased rate of hindbrain herniation (27.0% of respondents) and an improved rate of motor function (18.9% of respondents).Ranked last was the motivation to give parents a treatment option during gestation (18.2% of respondents).
When comparing postnatal surgery with open fetal surgery and fetoscopic surgery, respondents expected the risk of maternal complications to be the highest for open fetal surgery (89.2% of respondents), closely followed by fetoscopic surgery (79.9% of respondents; Table 4).Conventional postnatal closure was expected to have the lowest risk (92.5% of respondents).Open fetal surgery (70.6% of respondents) and fetoscopic surgery (66.8% of respondents) were expected to have the highest risks for perioper-ative fetal complications.No significant differences could be found in the expectations between countries providing fetal surgery and countries that do not.Furthermore, no significant differences could be identified between residents and neurosurgeons regarding barriers, facilitators, and expectations of fetal surgery (Fig. 3).
Table 5 gives an overview of respondent answers regarding the outcomes of open prenatal surgery.Among all respondents, 62.6% assumed that fetal surgery would lead to a higher risk of neonatal death, which contrasts with what the MOMS showed.
Table 6 gives an overview of respondent opinions regarding the position of fetal and fetoscopic surgery in the management of MMC.More than half of the respondents (54.4%) disagreed or strongly disagreed that open fetal surgery should be the standard of care.Similarly, more than half of the respondents (58.1%) disagreed or strongly disagreed that potentially damaging women as a result of fetal surgery is acceptable.Pediatric neurosurgeons were more likely than nonpediatric neurosurgeons to disagree on both of these items (p < 0.001 and p = 0.003, respectively).Reactions to the statement "denying potential improvement in the quality of life of MMC patients by not offering fetal surgery is acceptable" were more divided with almost a third (strongly) agreeing, a third being neutral, and almost a third (strongly) disagreeing.Slightly more than half (50.7%) of the respondents strongly disagreed or disagreed with the statement that providing fetoscopic surgery without performing a randomized controlled trial is acceptable.

Discussion
This study gives an overview of the current management of MMC and opinions on fetal surgery by neurosurgeons worldwide.Open fetal surgery (13.0%) and especially fetoscopic surgery (4.9%) are provided by only a minority of the centers in which the respondents are employed.Despite the well-documented decreased rates of shunt dependency and hindbrain herniation and improved rate of motor function in strictly selected pregnancies, only 16.9% of respondents agreed that open fetal surgery should be the standard of care.The most important objections to the procedure are a limited caseload available to become proficient in the technique, the risk of maternal complications, and the risk of fetal complications.The bind in the trade-off between maternal health and the health and quality of life of patients with MMC is illus-

Comparison With Other Studies
When the results of the MOMS were published in 2011, a paradigm shift was expected. 2However, such a shift was not reflected in our results: 23.7% of respondents never discuss the option of fetal surgery during counseling, and only 16.9% agree that open fetal surgery should be standard care.Moreover, only 13.0% of respondents offer open fetal surgery, while 4.9% offer fetoscopic surgery.An explanation for these low rates may be the generalizability of the MOMS, which included pregnancies carrying fetuses with MMC based on strict inclusion criteria.Results of fetal surgery for other pregnancies are less well-documented and warrant prospective studies.Another explanation may be the maternal comorbidity and the risk for future pregnancies.In the literature, a maternal complication rate of 6.2% for fetoscopic surgery and a rate of 20.9% for open fetal surgery have been reported. 29When taking into account only serious maternal complications such as those requiring ICU care or surgical intervention, risk rates of 1.7% and 4.5% have been reported for fetoscopic and open procedures, respectively.
Our results showed that 62.6% of the respondents assume that open fetal surgery carries a higher risk of neonatal death.Moreover, almost 40% wrongly assume a higher maternal mortality rate.These data may imply that a factor contributing to the nonoccurrence of the paradigm shift may be a lack of knowledge-on the one hand, a lack of knowledge because the outcomes reported in the literature may not be known among respondents, and on the other hand, because the results reported in the literature (e.g., MOMS) may not be generalizable to all pregnancies carrying fetuses with MMC.
Previous studies have shown that fetal surgery, especially fetoscopic surgery, is technically demanding and requires dedicated training. 12,13,20This may be reflected by the fact that only 4.9% of the departments surveyed provide fetoscopic surgery.A number of studies have been conduct-  ed to assess the relationship between hospital and surgeon volume and surgical mortality and morbidity. 7,24These studies have shown that a high surgeon volume results in superior outcomes for patients.Paradoxically, the majority of our respondents also agree that performing only one case of fetal surgery per month is acceptable, which mirrors the respondents' current practice of approximately 9.6 MMC repairs per year.
In light of the high case volume, the Maternal-Fetal Management Task Force released a position statement in which minimum criteria are recommended to ensure optimal maternal and fetal outcomes. 10In this statement, an initial experience of 5 cases is required, as is an adequate annual volume of cases to maintain proficiency.The exact number of cases meant by "adequate" is subject to debate, however.A recent meta-analysis on the learning curves of fetal surgery estimated that competence for standard hysterotomy is reached after 35 cases and that competence for minimally invasive procedures is reached after 56 cases. 20n this context, a smaller number of centers dedicated to fetal surgery, in which surgeons perform larger volumes of cases, is more desirable than a more disseminated offering of fetal surgery.A remedy toward achieving and maintaining proficiency may be adequate training models. 26Assisted nowadays by 3D-printed models, it is possible to create case-specific models to prepare for complex procedures such as fetoscopic surgery.Another debatable subject is the position of fetoscopic surgery.In a recent meta-analysis, outcomes of fetoscopic and open repair of MMC defects, published after MOMS, were analyzed and compared. 21A total of 11 studies were included, 4 retrospective and 7 prospective, comprising 436 patients.Fetoscopic MMC repair (percutaneous or laparotomy access combined) was associated with a lower rate of uterine dehiscence; however, this came with a higher rate of dehiscence or CSF leakage of the MMC defect, requiring postnatal revision.There were no betweengroup differences regarding fetal and postnatal mortality, rate of shunt placement, reversal of hindbrain herniation, motor response, gestational age before birth, chorioamniotic membrane separation, or placental abruption.Expectations of our respondents were similar with 89.2% expecting open fetal surgery to have the highest risk for maternal complications.Authors of another meta-analysis concluded similarly, suggesting that before performing a randomized study, endoscopic techniques should be improved first. 5o date, no randomized controlled studies have compared the two techniques of fetoscopic and open postnatal repair.Only a minority of our respondents, 21.9% (strongly) believe it is acceptable to provide prenatal fetoscopic surgery without performing a randomized controlled trial.Fetoscopic surgery faces multiple barriers to implementation, as indicated by the reluctance of neurosurgeons to offer open fetal repair for MMC with only 13.0% of centers in our respondent pool offering the procedure.

Study Strengths and Limitations
Some limitations of this study must be acknowledged.First, the response rate could be a limitation.The survey was mailed to 7671 functional email addresses of CNS members.Three hundred seventy-four responses were from CNS members leading to an estimated response rate of 4.9%.Of note, some of the email addresses may have belonged to medical students or other allied health professionals who may not have been eligible to fill out the survey.The importance of the response rate when judging the quality of cross-sectional research, however, is subject to debate. 17,23Given that 446 neurosurgeons with variable tenure and representing 58 countries responded to our survey, we expect a limited risk of nonresponse bias.However, we assume that MMC-dedicated centers, which also offer fetal surgery, may have been more inclined to respond.Another limitation may be the unfamiliarity of the respondents with either the care of patients with MMC or prenatal surgery.However, 81.0% stated that they treat patients with MMC and 91.0%worked at a department where either postnatal or prenatal surgery is provided.

Conclusions
The survey results show the diversity in the management of patients with MMC.In addition, significant diversity remains regarding fetal surgery for MMC closure.Despite the apparent benefits of open fetal surgery in selected pregnancies, only a minority of centers and providers offer this technique.As a more technically demanding technique that requires multidisciplinary effort with lesswell-established long-term outcomes, fetoscopic surgery may face similar limited implementation, although the surgery may pose fewer maternal risks than open fetal surgery.Centralization of prenatal treatment to tertiary care referral centers, as well as the use of sophisticated training models, may help to augment the most commonly cited objection to the implementation of prenatal closure, which is the overall limited caseload.

FIG. 1 .
FIG. 1.A geographic overview of the working locations of the respondents.Countries marked with a star have one or more respondents providing prenatal closure techniques.Created with mapchart.net©.CC BY-SA 4.0 (https://creativecommons.org/licenses/ by-sa/4.0/).

FIG. 2 .
FIG. 2.The most important objections to providing open fetal surgery for the treatment of MMC.The y-axis represents the percentage of respondents.The x-axis represents the rank chosen for each objection.

FIG. 3 .
FIG. 3. The most important reasons to provide open fetal surgery for MMC.The y-axis represents the percentage of respondents.The x-axis represents the rank chosen for each expectation.

TABLE 2 . Items discussed during counseling by pediatric versus nonpediatric neurosurgeon respondents
CONTINUED ON PAGE 5 »

TABLE 5 . Respondent knowledge about open prenatal surgery performed before 26 weeks of gestation
Answers in boldface type indicate results of the MOMS.

TABLE 6 . Overview of respondent opinions regarding the place of fetal surgery in the management of MMC
QOL = quality of life; RCT = randomized controlled trial.