Fetoscopic Repair of Isolated Fetal Spina Bifida

Spina bifida is a congenital anomaly that results from incomplete closure of the neural tube between 22 and 28 embryological days. Its incidence is approximately 2-4 cases per 10,000 births, and it is considered the most common congenital central nervous system anomaly that is compatible with life (CDC). Open spina bifida can present as a flat defect without a covering (myeloschisis), it may have a membranous covering (meningocele), or the fluid may be extruded into a fluid filled sac (myelomeningocele or MMC). Spina bifida can lead to lifelong sequelae that are the result of additional insult to the nervous system that occurs during fetal life as a consequence of the anomaly in the spinal cord. Downward displacement of the brain stem results in hindbrain herniation and the Chiari II malformation during fetal life leading to non-communicating hydrocephalus. Concurrently, intrauterine injury to exposed neural elements leads to neurologic dysfunction.

Despite improved care and technology, 2-year survival of affected individuals is 75%. The need for ventriculoperitoneal shunting for hydrocephaly is related to the level of the lesion and ranges between 88-97% for thoracolumbar lesions. Shunt placement in and of itself is associated with complications such as obstruction, infection, and displacement requiring repeated shunt revisions as early as the first year of life. The majority (75%) of patients with hydrocephaly have radiologic evidence of the Arnold-Chiari II malformation (hindbrain herniation, brain stem abnormalities, and small posterior fossa), which are associated with symptoms of apnea, swallowing difficulties, quadriparesis, balance issues, and coordination difficulties. The lesion level also correlates to the functional motor level; in general, the rate of being wheelchair-bound increases from 17% in sacral lesions to 90% of patients with a thoracic level lesions. Almost 90% of infants with spina bifida require intervention for a foot deformity to allow weight bearing activities. Bowel and urinary tract complications are common, and while children with spina bifida can achieve normal intelligence, they are at risk for neurocognitive and language difficulties that might impact school performance and the ability to live independently. The acquired disabilities tend to increase into adulthood and are attributed to a high rate of unexpected death. Overall, the most frequent form of spina bifida is MMC associated with hydrocephaly, lower limb paralysis, and bowel and bladder dysfunction.

The ultimate neurologic deficit that occurs with MMC that is established at birth is thought to originate from two mechanisms. First, there is an anatomic abnormality of a relatively normal spinal cord that then becomes secondarily damaged by the intrauterine environment through amniotic fluid exposure, direct trauma, hydrodynamic pressure, or a combination of these. This "two-hit hypothesis" is based on the observation that progressive neurologic damage develops in fetuses with MMC as gestation advances and results in irreversible neurologic damage at birth. The potential ability to ameliorate secondary damage caused by exposure to the in utero environment gave rise to the concept of fetal surgery for MMC repair.

The prenatal diagnosis of MMC with Chiari II malformation can be determined by ultrasound in almost all cases between 14-20 weeks gestation, and in the second trimester diagnosis is 97% sensitive and 100% specific. Accordingly, potential candidates for prenatal repair can be identified early in pregnancy leaving adequate time for detailed anatomic evaluation, genetic workup, and multidisciplinary patient counseling.

Because of the lifelong morbidity associated with the condition and the ability to accurately make a prenatal diagnosis of spina bifida, the idea of in-utero surgery to improve outcomes was conceived. Early animal studies and subsequent human pilot studies laid the groundwork for the Management of Myelomeningocele Study (MOMS trial). While the animal models supported the concept of the two-hit theory and the principle of improved neurologic function after in utero repair, these findings could not be directly extrapolated to human application. The National Institutes of Health (NIH) sponsored the multicenter randomized MOMS trial that compared outcomes between prenatal MMC repair with standard postnatal management. Prenatal MMC repair was performed between 19-25 6/7 weeks gestational age. The fetal repair involves a two to three layer closure similar to neonatal surgery. The neural placode is sharply dissected from the surrounding tissue. The dura and myofascial flaps are then re-approximated over the neural placode. A running suture is then used to close the skin. The coverage must be completely "water tight" to prevent the leakage of cerebrospinal fluid through the MMC defect that leads to hindbrain herniation in order to prevent amniotic fluid exposure which damages the neural tissues in the MMC defect. The uterus was closed in two layers (running closure and interrupted stay sutures) and then covered with an omental flap. Following prenatal MMC repair, patients remained near the fetal surgery center until delivery by cesarean section.

The study demonstrated that prenatal MMC repair was associated with a significantly lower rate of shunting, and hindbrain herniation and produced better motor outcomes. In the prenatal surgery group, functional motor level was better by two or more levels from anatomic level in 32% and better by one level in 11% compared with 12% and 9%, respectively.

The major risks of prenatal surgery for the fetus include chorioamniotic membrane separation (26% vs. 0%, p < 0.001), spontaneous rupture of membranes (46 vs. 8% p < 0.001), and spontaneous preterm labor with preterm birth (38 vs. 14%, p<0.001). This led to the lower gestational age at delivery in the prenatal surgery group of 34 weeks compared with 37 weeks in the postnatal surgery group (p < 0.001). Within the prenatal surgery group, 13% of patients delivered at a severely premature gestation of < 30 weeks and 33% at 30-34 weeks.

Prenatal MMC repair is associated with significant maternal risks, including pulmonary edema (6%) and blood transfusion at delivery (9%). Hysterotomy thinning was observed in 25% of women and uterine dehiscence and 1% of women. Moreover, women have a 14% risk for scar dehiscence in future pregnancies and invariably require delivery by cesarean section. The reason for the increased incidence of these complications is related to the nature of the open fetal procedure, which involves a multi-faceted invasive approach including maternal laparotomy, large hysterotomy with uterine edge stapling, and open fetal repair of the spina bifida defect that may involve manipulation and exposure of the fetus for a significant amount of time. Nevertheless, the MOMS trial demonstrated significant fetal and neonatal benefit. While maternal risks remain significant, prenatal MMC repair has been adopted as an accepted care standard across the United States.

Fetal endoscopic surgery has progressed rapidly over the past few decades and many fetal therapy centers are now able to perform a number of intricate procedures inside the uterus. Since fetoscopy offers a less invasive therapeutic option than open fetal surgery, there have been several efforts to develop this technique for MMC repair with the goal to duplicate the beneficial fetal effects while avoiding the significant maternal morbidity. Animal and human experimental experience with fetoscopic repair of MMC has been reported, showing the feasibility of covering the defect with a patch, sealant, or by full repair. These fetoscopic repairs are typically performed using at least two ports. Due to the complex surgical manipulations, particularly when patch closures are performed, operative times are long and are associated with significant obstetric morbidities.

A maneuver associated with improved ability to perform a fetoscopic repair is intrauterine insufflation with carbon dioxide. This provides a dry working area for the surgeon to perform the closure. The fetal therapy team at the Johns Hopkins Center for Fetal Therapy has previously utilized intrauterine carbon dioxide (CO2) insufflation in situations where a dry surgical environment was required. Most recently a two port-technique for fetoscopic MMC repair under CO2 insufflation was described by the Baylor College of Medicine/Texas Children's Fetal Center using the externalized approach. This technique employs a laparotomy to exteriorize the uterus, which can then be positioned for access with two surgical ports regardless of the placental location. After CO2 insufflation and fetal anesthesia is administered, the MMC repair is performed after sharp dissection of the placode using a mattress suture. This approach is designed to decrease the maternal obstetric risks while preserving the fetal benefits.

The technique employs low-pressure uterine CO2 distention at 8-12 mmHg. In addition, significantly quicker neural tube repair is possible because of improved access to the fetus, ability to manipulate the fetus into the required position, and superior port placement resulting from the exteriorized maternal uterus. As a result only two ports are required and these can be sutured into the uterus allowing a closed seal and minimizing gas leakage. Finally, recent advances in small diameter surgical instruments (Storz 1.5 - 3mm surgical sets) allow a full surgical repair to be performed via a fetoscopic approach.

The purpose of the current study is to evaluate the feasibility of performing fetoscopic spina bifida repair at Johns Hopkins Hospital and the fetal and maternal outcomes following this approach.