- 224 Downloads
The first description of open maternal-fetal surgery for correction of anatomic anomalies by Harrison was published three decades ago (Adzick 2003). At that time, the diagnostic and surgical tools for prenatal treatment of the fetus were just being developed, and the concept of the fetus as a patient was the subject of philosophical and ethical debate. Since then, great progress has been made in the ability to diagnose fetal abnormalities, predict their outcome, and perform surgical interventions when appropriate. The concept of the fetus as a patient has become a standard of care, and the ethical framework for maternal-fetal intervention is well developed (Adzick, Semin Fetal Neonatal Med 15(1):1–8, 2010). While application of open fetal surgery has remained limited to a relatively small number of highly selected fetuses and is practiced in only a few centers, the development of this field has accelerated technological progress in prenatal diagnosis and intervention, led to improved understanding of the pathophysiology and natural history of candidate disorders, allowed comprehensive counseling of parents in centers with focused expertise in fetal anomalies, and driven the evolution of less invasive therapeutic approaches. The purpose of this chapter is to describe the current status of fetal surgical intervention and to speculate regarding future developments in this rapidly evolving field.
KeywordsFetal surgery Fetoscopy Ex utero intrapartum therapy Congenital cystic adenomatous malformation Congenital diaphragmatic hernia Myelomeningocele Sacrococcygeal teratoma Twin-to-twin transfusion syndrome
Fetal surgery is a specialty born of clinical necessity. Observations of fetuses with congenital anomalies and neonates with irreversible end-organ damage led to the compelling rationale that correction in utero might arrest progression of disease or even reverse the pathophysiology and restore normal development. First, research in animal models was needed to study the development and progression of specific fetal defects. Those early endeavors not only validated the initial hypothesis but also lead to innovative techniques and technologies to further clinical application.
The earliest report of successful therapeutic intervention on the fetal patient was intraperitoneal transfusion of a hydropic fetus for Rh disease by Sir Liley in 1963, and this represents the first acknowledgment of the fetus as a patient (Jancelewicz and Harrison 2009). The modern era of fetal surgery was conceived and developed by Michael R. Harrison and colleagues at the University of California, San Francisco (UCSF), during the late 1970s and 1980s. The same group performed the first open fetal surgery for an anatomic anomaly creating bilateral ureterostomies in a fetus with congenital bilateral hydronephrosis due to urinary tract obstruction (Harrison et al. 1982).
Anatomic anomalies currently treated by fetal therapy
Rationale for in utero therapy
Extrinsic or intrinsic airway compression
Stabilization of airway and circulatory support before interruption of uteroplacental gas exchange
Ex utero intrapartum therapy (EXIT)
Congenital lung lesion
Reversal of pulmonary hypoplasia and cardiac failure
Open surgery: lobectomy; fetoscopic shunting of macrocystic lesions
Congenital diaphragmatic hernia (CDH)
Reversal of pulmonary hypoplasia and pulmonary hypertension
Fetoscopic tracheal balloon occlusion (FTO)
Protection of exposed spinal cord and cessation of cerebrospinal fluid leakage; prevention or reversal of hindbrain herniation and hydrocephalus
Open surgery: closure of defect
Sacrococcygeal teratoma (SCT)
Reversal of steal phenomenon and high-output cardiac failure; prevention of polyhydramnios
Open surgery: tumor debulking
Twin-to-twin transfusion syndrome (TTTS)
Normalization of inter-twin transfusion; reversal of cardiac failure
Fetoscopic laser photocoagulation
Imaging and Diagnostics
Current progress in fetal diagnosis, intervention, and treatment would not have been possible without the prerequisite advances in laboratory diagnostic and imaging technologies. Fetal ultrasound was first described in the late 1960s and remains the primary imaging modality for prenatal screening and diagnosis with its proven utility, relatively low cost, and widespread availability (Hopkins and Feldstein 2009). Further, ultrasonography is advantageous due to its multiplanar capability, Doppler flow depictions, high spatial resolution, and real-time assessment. Limitations of ultrasonography include a relatively small field of view, beam attenuation by maternal adipose tissue, poor image quality in oligohydramnios, poor acoustic access to the fetal head when it lies deeper in the pelvis, and limited visualization of the posterior fossa due to calvarial calcification later in gestation. When fetal intervention is a consideration, diagnostic certainty supported by accurate and specific details is paramount to the plan of care, and magnetic resonance imaging (MRI) is a useful adjunct and may be required for specific anomalies (Bulas 2007). The advantages of MRI include a larger field of view, superior soft-tissue contrast, more precise volumetric measurements, and greater accuracy in demonstrating intracranial abnormalities. However, long-term safety after exposure to high-field MRI has not been demonstrated, and there are concerns for effects on embryogenesis, chromosomal structure, and fetal development. As a precaution, MRI is not recommended during the first trimester. Gadolinium contrast crosses the placenta and has been found in animal studies to be associated with growth retardation. Therefore, gadolinium contrast is not recommended in pregnant patients.
As for prenatal laboratory diagnostics, Geaghan provided a thorough review of current techniques (Geaghan 2012) including but not limited to amniocentesis, chorionic villus sampling (CVS), cordocentesis, and maternal blood sampling for fetal products.
Principles of Maternal-Fetal Surgery
Indications and Contraindications
The prerequisites for consideration of open fetal surgery include (1) an accurate prenatal diagnosis and exclusion of associated anomalies; (2) presence of a correctable lesion that, if left untreated, will lead to fetal demise or severe irreversible organ damage before birth; (3) well-defined natural history of the disorder allowing selection of fetuses that will benefit from prenatal intervention; and (4) technical feasibility of fetal surgery with an acceptable risk-to-benefit ratio for both mother and fetus (Chervenak and McCollough 2007). Contraindications to fetal surgery include complex chromosomal or associated anatomic abnormalities in the fetus, maternal risk factors including incompetent cervix, placentomegaly, maternal mirror syndrome, morbid obesity, any serious comorbid conditions, including prohibitive psychiatric or psychosocial disorders, drug or alcohol abuse, and heavy smoking.
Any patient carrying a fetus diagnosed with an anomaly that may require fetal intervention should be referred to a fetal treatment center for a multidisciplinary evaluation, including detailed ultrasonography, fetal MRI, and fetal echocardiography. In addition, fetal karyotype, rapid fluorescent in situ hybridization analysis for common aneuploidies, and hybridization arrays should be performed to rule out major chromosomal abnormalities and common genetic defects. Thorough counseling for patients based on this extensive workup is crucial and includes a detailed nondirective discussion of all available options for the pregnancy and the risks and benefits of each, including nonoperative management, termination, and palliative care. If open fetal surgery is an option, the mother should be counseled on the risks and consequences, including the future risk of uterine rupture and the need for preemptive cesarean delivery in the current and all future pregnancies (Wilson et al. 2004).
Anesthesia for maternal-fetal surgery is uniquely challenging because two patients are being cared for at once. Patients should be admitted before any planned open fetal procedure for monitoring and initiation of tocolysis. Because of the risk for chorioamnionitis, intravenous antibiotics are given before incision. A type and screen are sufficient for most minimally invasive procedures, but open procedures require available cross matched blood for the mother and warm type O-negative blood for the fetus which can be cross matched with the maternal sample to avoid reaction with maternal antibodies that cross the placenta.
Anesthesia for open fetal cases is a combination of epidural anesthesia and deep inhalational general anesthesia. Anesthesia is initiated with placement of an epidural catheter to provide both intraoperative and postoperative pain management. General anesthesia is then induced with inhalational agents, with a goal of minimum alveolar concentration of 2–2.5 to provide adequate uterine relaxation. Maternal monitoring includes blood pressure monitoring via radial artery catheter and blood pressure cuff, a Foley catheter, continuous electrocardiographic monitoring, and pulse oximetry. Multiple large-bore intravenous catheters are inserted for access. Sequential compression devices are used to prevent deep venous thrombosis, and fluid management is aimed at euvolemia to prevent maternal postoperative noncardiac pulmonary edema.
Positioning and Draping
The patient is positioned supine with left lateral tilt obtained by placing a roll under the right side to minimize aortocaval compression from the gravid uterus and to increase venous return. The skin is prepped from the mid-thorax to the mid-thigh, and the operative field is squared off with sterile towels and covered with a fenestrated and pocketed drape.
Incision and Exposure
The initial incision is generally a low transverse abdominal incision. Precise mapping of the placenta is paramount because placental position determines the type of fascial opening, as well as hysterotomy placement. If the placenta is posterior, subcutaneous flaps are raised and the fascia divided in the midline from the umbilicus to the symphysis pubis, but if it is anterior, the muscle and fascia must be divided transversely to allow for anterior rotation of the uterus and a posterior hysterotomy. Once the uterus is exposed, a ring retractor is positioned for exposure.
Before the hysterotomy is created, the uterus is palpated to evaluate for sufficient relaxation. Fetal and placental position are then confirmed with ultrasound. Electrocautery is used under ultrasound guidance to map placental margins on the uterine surface, and a safe site is identified that avoids the placenta and uterine vasculature. The lower segment of the uterus is avoided due to an increased postoperative risk of preterm labor, amniotic fluid leak, and chorioamnionitis.
Closure of the gravid uterus requires adequate strength to prevent rupture and amniotic fluid leak, but must do not contribute to infertility in the future. After return of the fetus to the amniotic space, the closure is performed in two layers. Full-thickness double-armed absorbable stay sutures placed approximately 2 cm apart and 2 cm away from the staple line are placed first and held on traction, while a running second suture line is placed just outside the staple line through the myometrium and membranes (Fig. 1d). Warmed lactated Ringer’s solution is infused to restore amniotic fluid volume to baseline, and intra-amniotic antibiotic is administered before tying the running layer of suture. The interrupted stay sutures are then tied, and the uterine closure is buttressed with an omental flap. The maternal laparotomy is closed in multiple layers. Skin is closed using a subcuticular running absorbable suture, and a transparent dressing is applied in order to allow for continued fetal ultrasound monitoring postoperatively.
Because preterm labor is a relatively common complication, tocolysis is continued postoperatively. Pain management is continued using the epidural catheter placed preoperatively, which prevents uterine irritability once the inhaled anesthetic has worn off. Daily fetal echocardiography is performed to evaluate for evidence of indomethacin toxicity, which may manifest as ductal constriction, tricuspid regurgitation, or oligohydramnios. A tocodynamometer is used to monitor uterine activity, and the fetal heart rate is followed for signs of distress. Ultrasounds are performed daily while the patient is in the hospital to assess amniotic fluid and membrane status, anatomic status, and fetal movement.
Most maternal-fetal surgery patients are discharged by postoperative day 4 but should remain on modified bed rest for 2 weeks following discharge. Patients can return to moderate activity after that if no signs of uterine irritability are present. Patients should return for ultrasounds twice per week until delivery. At 36 weeks gestation, an amniocentesis is performed to assess fetal lung maturity, and the fetus is delivered via cesarean section once the lungs are mature.
While no maternal deaths have been reported following maternal-fetal surgery to date, these procedures are associated with significant short-term morbidity. In a retrospective study of 178 women who underwent open, fetoscopic, or percutaneous ultrasound-guided procedures, the most commonly observed complications included preterm premature rupture of membranes, preterm labor, and preterm delivery. Additional complications that have been observed include chorion amnion membrane separation, chorioamnionitis, placental abruption, pulmonary edema, and bleeding requiring transfusion. The overall complication rate is significantly lower for percutaneous procedures. Subsequent pregnancies after open maternal-fetal surgery are at risk for uterine dehiscence and rupture if labor is allowed to occur, but fertility is not affected.
When first developed in the 1970s, fetoscopy functioned as a diagnostic tool but has since become a tool for minimally invasive fetal interventions with the development of more advanced camera equipment and endoscopic devices. Fetoscopy has similar complications to open fetal surgery, including bleeding, preterm premature rupture of membranes, preterm labor with delivery <32 weeks, chorioamniotic separation, preterm delivery, and chorioamnionitis. The risk of fetoscopic complications correlates with the size of the fetoscope, the number of trocar sites, and the length of the procedure. For simple procedures, the risk of preterm labor is lower than open fetal surgery. Most of these procedures can be performed with only local anesthesia. However, any procedure that would potentially cause fetal pain should include intramuscular injection of an opioid and a paralytic agent for fetal anesthesia.
Ex Utero Intrapartum Therapy (EXIT)
The EXIT procedure was developed to allow removal of tracheal clips and establishment of an airway at the time of delivery after prenatal tracheal occlusion had been performed in severe cases of CDH. However, the indications for EXIT have expanded to include any case in which difficulty in obtaining an airway is anticipated, including congenital high airway obstruction syndrome (CHAOS), giant anterior neck masses including cervical teratomas, pharyngeal tumors, mediastinal tumors, and hypoplastic craniofacial syndrome, or cases in which a thoracic mass will require resection in order to ventilate the patient as in giant thoracic masses with dramatic mediastinal shift (large solid CPAMs) (Bouchard et al. 2002).
Similar to open maternal-fetal surgery, the patient is positioned supine, with left lateral tilt to maximize blood flow to the uterus and placenta as well as venous return. Inhalational anesthesia is utilized to achieve maximal uterine relaxation, and intramuscular narcotic and paralytic agents are given to prevent fetal discomfort and movement. As with open maternal-fetal surgery, ultrasound is used to map the placental edges, and the uterus is opened with the same absorbable stapler used in open fetal surgery to prevent blood loss and control the membranes. In cases with severe polyhydramnios or large cystic masses, amnioreduction or cyst aspiration may be required before hysterotomy is performed. A peripheral intravenous line is obtained on the fetus, a pulse oximeter is applied to the exposed fetal hand, and continuous fetal transthoracic echocardiography is utilized to assess fetal cardiac function and volume status. Warmed lactated Ringer’s solution is infused to maintain uterine volume and fetal temperature. The placental circulation is kept intact while an airway is established. Intubation may be performed either ante- or retrograde; the later accomplished using the Seldinger technique via a limited neck dissection, keeping in mind that tracheal anatomy may be significantly distorted in cases of giant fetal neck masses, and the carina may be superiorly displaced. Rigid bronchoscopy and/or tracheostomy may be used in some cases, and decompression or debulking may be required in order to obtain an airway (Moldenhauer 2013).
A successful EXIT procedure relies on adequate uterine relaxation to maintain uteroplacental blood flow and prevent placental separation. However, this relaxation can lead to maternal postoperative bleeding. Additional maternal risks of EXIT include uterine rupture or scar dehiscence in a subsequent pregnancy and wound infection. The goal is to have the fetus off placental support within 60 min, although procedures up to 2.5 h have been reported. Risks to the fetus during EXIT include hypoxic brain injury, bradycardia, hemorrhage, and death secondary to cord compression, loss of myometrial relaxation, and placental abruption, all of which cause inadequate uteroplacental gas exchange.
Anatomic Anomalies Currently Managed by Fetal Surgery
Congenital Bronchopulmonary Malformations
Congenital bronchopulmonary malformations represent a continuum of abnormalities of the bronchopulmonary unit for which classification remains in evolution. They include a spectrum of disease entities ranging from congenital pulmonary airway malformation (CPAM – formerly congenital cystic adenomatoid malformation or CCAM) to bronchopulmonary sequestration (BPS), with an intermediate subset of “hybrid” lesions which exhibit characteristics of both. While they differ considerably in their pathologic features, from a fetal perspective, these lesions are usually diagnosed by ultrasound as space-occupying cystic and/or solid lesions. There are no known associated anomalies with CPAM, whereas approximately 10% of fetuses with intralobar and 50% with extralobar BPS have associated anomalies, including tracheoesophageal fistula, congenital heart defects, CDH, aneuploidy, or foregut duplication.
BPS is a mass of nonfunctioning lung tissue of which the defining developmental feature is the absence of a communicating bronchus. These lesions have systemic arterial supply, which most often arises from the thoracic aorta but can arise from any systemic source. An extralobar BPS has its own investing pleura, no communication with the native bronchial tree, and usually systemic venous drainage although pulmonary venous drainage is occasionally observed. On the other hand, an intralobar BPS shares visceral pleural investment with the normal lung and generally has pulmonary venous drainage. The lesions may be aerated after birth due to tracking of air through the pores of Kohn and, thus, in contrast to extralobar BPS, are subject to mucostasis and infection.
Fetal BPS and CCAM lesions range in severity from small with minimal clinical significance to rapidly growing masses that cause mediastinal shift and cardiac failure related to tamponade physiology with secondary fetal hydrops. Fetal hydrops and placentomegaly from these lesions can lead to placental vasoreactivity, causing development of the maternal mirror syndrome, a condition where the mother develops progressive preeclamptic symptoms, which “mirror” fetal pathophysiology.
Diagnosis of cystic lung lesions is based on identification of an echogenic mass in the fetal chest (Fig. 2) that may be solid, cystic, or mixed. Doppler ultrasound will show either systemic (BPS), pulmonary (CCAM), or both (hybrid lesions) arterial blood supply. The differential diagnosis of these lesions includes peripheral bronchial atresia, bronchogenic or neurenteric cysts, congenital lobar emphysema, and CDH. Ultrasound of CCAMs may show mediastinal shift away from large lesions or polyhydramnios secondary to esophageal compression. Hydrops may be evident in severely affected fetuses. Lymphatic and venous congestion may lead to pleural effusion, which could cause tension hydrothorax and secondary hydrops.
Experience with serial imaging of large numbers of fetuses with CPAM has clarified the pre- and perinatal natural history of this anomaly. There is a typical pattern in CPAM growth consisting of a period of growth relative to the size of the fetus until approximately 26 weeks gestation when growth plateaus. After 28 weeks the CPAM typically gets smaller relative to the size of the fetus as measured by the CPAM volume ratio or CVR. The CVR is calculated by dividing CPAM volume (length × height × width × 0.52) by the head circumference. In addition, the CVR has proven on retrospective and prospective assessment to be the most useful predictor of the evolution of hydrops. Presentation with a CVR of ≤1.6 in a CPAM without a dominant cyst predicts a less than 3% risk of developing hydrops. If the CVR is >1.6, the risk of hydrops is around 75%. CVR has proved very useful in counseling patients, determining the intensity of serial follow-up, and choosing which patients to preemptively treat with steroids. Other parameters such a mass-thorax ratio, cystic predominance of the lesion, and eventration of the diaphragm, while associated with large lesions, do not add independent predictive value to the CVR.
Open fetal surgical resections for microcystic CCAM are associated with 60% survival with most patients enjoying a good quality of life. Thoracoamniotic shunt placements for macrocystic CCAM have been reported to decrease CCAM mass volumes by an average of 50% and up to 80% and are associated with approximately 75% survival.
Congenital Diaphragmatic Hernia
Congenital diaphragmatic hernia (CDH) is a developmental defect in the diaphragm, which leads to herniation of abdominal viscera into the chest. CDH affects 1 in 3000 live births and is most often sporadic, although familial cases have been reported. CDH is often syndromic; 25–57% of live born cases and 95% of stillborn fetal cases occur with associated abnormalities. These associated anomalies include hydronephrosis, congenital heart defects, renal agenesis, extralobar sequestrations, and neurologic defects including hydrocephalus, spina bifida, and anencephaly. Of prenatally diagnosed cases, 10–20% of CDH cases are associated with chromosomal abnormalities including trisomies 13, 18, and 21.
The diaphragmatic defect seen in CDH is the result of failure of the foramen of Bochdalek to close between 8 and 10 weeks of gestation. The pathophysiology of CDH consists of fixed pulmonary and vascular hypoplasia and reversible pulmonary vascular reactivity. The herniation of abdominal contents occurs at a critical phase of lung development when branching morphogenesis generates the normal bronchial and arterial tree. The resultant pulmonary hypoplasia includes varying degrees of reduced airway branching, alveolar structures, and vascular components. This leads to decreased lung surface area for gas exchange as well as a fixed increase in pulmonary vascular resistance. The pulmonary vasculature is also morphologically abnormal, with hypermuscular peripheral pulmonary arteries that have a thickened media. This causes increased pulmonary vasoreactivity and pulmonary hypertension. This resulting pulmonary hypertension leads to persistence of the fetal circulation, with shunting through the ductus arteriosus or foramen ovale, which then causes acidosis and hypoxemia.
The severity of CDH is related to the timing of herniation as well as the volume occupied by the herniated abdominal viscera in the thoracic cavity. If herniation occurs after lung development is nearly complete, the manifestations of the disease are much less severe, and a better outcome is seen. If, however, herniation occurs earlier in development, severe lung hypoplasia occurs, leading to a poorer prognosis. CDH therefore can be thought of as a spectrum of disease, ranging from mildly affected infants with relatively normal lungs to those with such severe hypoplasia that survival is unlikely.
Because CDH has a wide range in severity and a high frequency of associated anomalies, a complete prognostic assessment is critical (Hedrick 2013). This includes high-resolution ultrasound, fetal MRI, echocardiography, and genetic testing, all between 20 and 24 weeks gestation. This time frame allows for complete counseling for families, with the option for elective termination. The extreme importance of accurate counseling has led to investigation of factors predictive of poor outcome in CDH fetuses. CDH with associated major anomalies has a very poor prognosis. The only reports of CDH survivors with congenital heart disease (CHD) have a combination of relatively mild CDH and cardiac biventricular anatomy. Mortality associated with severe CDH and univentricular CHD nears 100%, and comfort care should be offered. Poor outcomes are also associated with familial CDH, bilateral CDH, CDH associated with specific genetic abnormalities, and syndromic CHD.
Liver herniation has historically been the most important poor prognostic indicator in CDH and can be assessed by ultrasound or MRI. In left-sided CDH, the presence of liver in the chest is associated with a very large defect, indicative of early herniation of viscera, causing severe pulmonary hypoplasia. A recent study showed mortality of 65% when the liver is up versus 7% when the liver is below the diaphragm. In addition, liver position proved to be predictive of the need for postnatal extracorporeal membrane oxygenation (ECMO), with 80% of liver up patients requiring ECMO, versus 25% of liver down patients.
In addition to herniation of the liver, various indirect measurements of lung volume have been developed with prognostic relevance to CDH. The ratio between right lung area (measured at the level of the four-chamber heart view) and head circumference (LHR) can be measured by ultrasound and has been validated as a prognostic indicator when measured between 22 and 24 weeks gestation. The clinical utility of LHR is controversial, as the measurements are subjective and widely dependent on the skill and experience of the sonographer. The most widely used lung measurement to predict morbidity and mortality is the observed to expected lung area to head circumference ratio (O/E LHR), which is measured by ultrasound or MRI. However, many CDH patients who have what appears to be an adequate lung volume for survival have significant morbidity and mortality from the disease due to pulmonary hypertension. Therefore, it is unlikely that prenatal lung volume estimations will ever provide complete prognostic accuracy due to the poor correlation between lung volume and pulmonary vascular bed reactivity.
Treatment of CDH
Prenatal management of CDH begins with thorough counseling, which relies heavily on an accurate diagnosis. It is paramount that the family understands the severity of CDH and the possible pre- and postnatal events that accompany it. The potential for poor outcome in a severe case of CDH, including death and severe pulmonary, gastrointestinal, and neurologic morbidity, should be discussed. The standard prenatal management for CDH is expectant, with ultrasound screening for prenatal complications. The majority of pregnancies with isolated CDH deliver at term, with a 3–8% stillbirth rate. CDH infants with polyhydramnios due to kinking of the gastroesophageal junction are at increased risk of preterm labor. Prematurity and its associated pulmonary insufficiency are often lethal when combined with the pulmonary hypoplasia seen in severe CDH. Ultrasound is recommended once a month up to 32 weeks gestation and then weekly to screen for polyhydramnios. The current algorithm for management of fetuses with CDH is shown (Fig. 5b).
The first attempted fetal intervention for CDH involved a patch repair of the defect. However, fetuses with liver up did not tolerate this intervention due to kinking of the umbilical vein, which led to intrauterine demise. In addition, there was no significant difference in survival for liver down-treated CDH fetuses repaired in utero when compared with postnatal repairs. Because of these limitations, open fetal repair was abandoned (Harrison et al. 1997).
Tracheal occlusion (TO) (Deprest et al. 2010) is a more recent fetal intervention of interest for CDH and treatment of pulmonary hypertension. The theory behind TO for CDH is that fetal lungs are net producers of lung fluid and that lung growth is related to airway fluid pressure, normally regulated by laryngeal mechanisms. It has been shown in animal models that shunting fluid from the lungs to the amniotic space can induce pulmonary hypoplasia but that fetal lungs undergo hyperplastic growth when the trachea is occluded. Accelerated lung growth and improved pulmonary function have been shown in the rat nitrofen and fetal lamb models of TO in CDH. However, clinical trials for TO using open and fetoscopic approaches have shown mixed results, including a prospective trial performed at Children’s Hospital of Philadelphia (Flake et al. 2000) showing that neonates with CDH treated with TO had severe respiratory compromise, even when lung growth had occurred. A randomized, controlled trial of fetoscopic TO from UCSF failed to show benefit. More recently, Jan Deprest et al. (2011) along with the Eurofetus study group have applied a minimally invasive method for TO using a deployable balloon inserted through a single small trocar. The initial reported results are promising, and a multicenter randomized controlled trial in North America and Europe known as the Tracheal Occlusion to Accelerate Lung Growth trial has recently begun and will evaluate the efficacy of this technique. At the present time the efficacy of TO for CDH is unproven, and there is potential for harm using this technique. It should only be done in the context of a well-designed clinical trial to establish efficacy prior to further clinical dissemination.
Outcomes in CDH
Currently, survival for infants born with CDH at a tertiary center is 70–92%, which represents an improvement in survival relative to several decades ago. However, it is important to note with any discussion of CDH survival that comparisons can only be made between patients that are accurately stratified for severity. Improved survival is credited to a shift from early surgical intervention and aggressive ventilatory management to delayed surgery and parenchymal sparring strategies such as permissive hypercapnia and early ECMO if ventilatory criteria are exceeded. These numbers do not take into account cases of CDH that die outside a tertiary center or fetal loss due to abortion or stillbirth. Transport of infants with CDH is associated with worse survival than infants who are born at a tertiary center.
Morbidity for CDH survivors includes respiratory, musculoskeletal, nutritional, gastrointestinal, and neurological complications. The CHOP Pulmonary Hypoplasia Program has prospectively evaluated over 300 CDH survivors. Of the 41 CDH survivors initially studied, 90% were found to have abnormal muscle tone at 6 months and 51% at 24 months. Many CDH survivors suffer from diminished neurocognitive and language skills, and the risk of autism significantly increased (Danzer et al. 2016).The high incidence of morbidity combined with the increasing survival of CDH patients to discharge creates the prerogative for ongoing coordinated care for these patients.
Myelomeningocele (MMC) occurs in approximately 1 in every 3000 live births and remains one of the most common congenital defects despite widespread appreciation of the preventative effects of folic acid supplementation. This condition is characterized by a defect in the vertebral arches allowing protrusion of the meninges and neural elements with devastating neurologic consequences including paralysis of the lower extremities, developmental delay, and incontinence of bowel and bladder. MMC represents the first application of fetal surgery to a nonlethal disorder, culminating in the recent publication of the Management of Myelomeningocele Study (MOMS), which demonstrated a clear advantage of prenatal closure of MMC compared to standard postnatal treatment (Adzick et al. 2011; Adzick 2013).
Pathophysiology and Natural History
The conceptualization and validation of the “two-hit” hypothesis were a critical step in the consideration of MMC as a compelling target disorder for fetal therapy, despite its nonlethal nature. The first “hit” is the primary failure of neural tube closure, allowing for the resultant second “hit,” which is exposure of the neural elements to amniotic fluid and mechanical trauma within the intrauterine environment. There is a body of clinical and experimental evidence supporting the concept that the majority of the neural damage is related to the second hit, creating the compelling rationale for fetal surgical closure. The fetal lamb MMC model was most influential in supporting a clinical trial of prenatal MMC closure by confirming that amniotic fluid exposure of the exposed neural elements resulted in severe neural damage which could be prevented by prenatal closure of the defect (Meuli et al. 1995). In addition to the open neural defect, almost all fetuses with MMC display a constellation of neuroanatomic abnormalities referred to as the Arnold-Chiari II malformation, characterized by descent of the posterior fossa contents through the foramen magnum, with resultant hindbrain herniation, inferior displacement of the cerebellar vermis, and elongation and kinking of the medulla. The hindbrain herniation impairs normal circulation of cerebral spinal fluid and results in development of hydrocephalus requiring shunt placement in 80–90% of cases. Almost half of these patients experience shunt complications, including failure secondary to obstruction or infection within the first year. This contributes significantly to the morbidity and mortality of MMC as well as the cognitive deficit. Although 70% of postnatally repaired MMC patients have an IQ higher than 80, only half are able to live independently as adults, even with adapted accommodations.
The algorithm for treatment of MMC is shown (Fig. 6c). Fetal MMC repair is offered to patients based on the inclusion criteria established in the MOMS trial, including singleton pregnancy with an MMC at level T1 through S1, Arnold-Chiari II malformation, gestational age 19 to 25 weeks, and normal karyotype without coexisting severe anomalies.
The operative procedure begins with a low transverse laparotomy, followed by creation of a hysterotomy as described above. The fetus is positioned to expose the MMC lesion. Continuous intraoperative fetal echocardiographic monitoring is critical. Fetal anesthesia is provided by the maternal inhalational anesthetic, and a narcotic dose is delivered intramuscularly to the fetus. The cystic membrane of the MMC is excised and the spinal cord untethered. The dura is reflected over the defect and closed with a running suture, followed by the paraspinal myofascial flaps, and then the skin. If the skin cannot be closed primarily, an acellular dermal graft is used to assist with the closure. Cesarean delivery is mandated for this and all subsequent pregnancies.
The MOMs trial was powered to recruit 200 participants but was halted after randomization of 183 patients when a planned interim analysis demonstrated clear benefit for prenatal surgery (Adzick et al. 2011). The fetal surgery group showed significant reduction in rates of shunt placement at 1 year (40% versus 82%) and improvement in neuromotor function by 30 months of age, including the ability to walk without orthotics (42% vs. 21%). The degree and presence of hindbrain herniation were also improved, with no hindbrain herniation in 36% of fetal surgery patients and 4% of postnatal surgery patients and severe hindbrain herniation in 6% of fetal surgery patients and 22% of postnatal surgery patients. The benefits of fetal repair outweighed the complications related to prematurity and the maternal morbidity seen in the study (Golombeck et al. 2006).
Sacrococcygeal teratoma (SCT) is the most common solid tumor in the neonate with an incidence of 1 in 40,000 births and a female-to-male ratio of 4:1. These tumors arise from the primitive streak and are composed of elements from all three germ layers. The American Academy of Pediatrics Surgical Section classifies SCTs according to their relation to the pelvis: type I tumors are external, with a small presacral component, type II tumors are predominantly external with intrapelvic extension, type III tumors are predominantly internal with intrapelvic and intra-abdominal extension and a small external component, and type IV tumors are entirely presacral without external or intrapelvic extensions.
Pathophysiology and Natural History
SCTs are predominantly benign, though they have malignant potential. The majority of patients diagnosed late in gestation or postnatally do well after complete resection, which includes complete removal of the coccyx to prevent recurrence. The mortality rate for a prenatally diagnosed SCT ranges from 30% to 50%. This high mortality is attributed to a variety of factors. These prenatally diagnosed tumors are often large, and mass effect can lead to maternal-obstetric complications and preterm labor with associated fetal demise. More acutely, SCTs can hemorrhage internally causing rapid enlargement of the tumor, leading to fetal anemia. SCTs can also rupture into the amniotic cavity, resulting in sudden death. Arteriovenous shunting and the associated vascular steal phenomenon can lead to high-output cardiac failure, placentomegaly, and fetal hydrops. Fetal mortality approaches 100% once these latter processes develop. Prenatal indicators of poor prognosis include tumor size, rate of growth, predominantly solid composition, high vascularity, signs of high-output cardiac failure, placentomegaly, hydrops, and the occurrence of maternal complications.
The evolution of high-output physiology and secondary hydrops in a fetus with SCT is nearly always associated with fetal demise and supports the rationale for performing open fetal surgery. Fetuses with large type I tumors exhibiting clear evidence of early hydrops related to tumor flow prior to 28 weeks gestation are candidates for open fetal surgery with debulking of the tumor. Fetal intervention aims to prevent progression of vascular steal phenomenon and high-output physiology. If hydrops or placentomegaly should develop after 28 weeks, early delivery with debulking is recommended. This allows stabilization of the critically ill newborn in the neonatal intensive care unit (NICU) prior to definitive resection.
The hysterotomy site is chosen to allow exteriorization of the tumor and the caudal end of the fetus. However, the umbilical cord is at risk for compression against the rim of the hysterotomy in this position, and the fetus should be continuously monitored for signs of cord compression. Care must be taken to keep the remainder of the fetus within the amniotic sac; inadvertent delivery of the whole fetus can lead to uterine contraction, inability to place the fetus back in the amniotic sac, and preterm labor.
Once the tumor is exteriorized, a Hegar dilator should be placed in the rectum to delineate anatomy, and the skin around the anorectal sphincter is incised. Fetal skin around the base of the tumor is then incised, controlling the large subcutaneous veins. A tourniquet is applied around the base of the tumor where the skin has been incised to restrict blood flow to the tumor. A handheld harmonic scalpel is then used to divide the tumor at its base, using suture ligation for larger vessels. Any intrapelvic component of the tumor should be left as well as the coccyx, to be excised at the time of definitive resection postnatally. Once the tumor bed is hemostatic, the fetus can be returned to the amniotic cavity and the hysterotomy closed, as described in previous sections.
Postoperatively, because of the risk of maternal mirror syndrome, maternal fluid balance should be closely monitored, and fetal echocardiography should also be performed frequently to follow resolution of the hydrops and placentomegaly.
Fetal SCT represents the most challenging of the anomalies treated by open fetal surgery. The derangement of fetal and maternal physiology results in a high rate of preterm labor with relatively short intervals between fetal intervention and delivery. For appropriately selected fetuses, survival rates of 50–70% can be expected; however, quality of life is variable with a high potential for severe morbidity. Parents should be fully informed of both the negative and positive outcomes after fetal intervention, and fetal surgery should only be undertaken when the specific indication of high-output cardiac failure is present with evidence of impending cardiac decompensation and early hydrops. After 27–28 weeks, preemptive cesarean delivery at the first sign of fetal or maternal decompensation is preferable to fetal surgery with immediate debulking of the tumor and transfer to the NICU (Roybal et al. 2011). An algorithm for management of the fetus with SCT is shown in Fig. 7b.
Twin-to-Twin Transfusion Syndrome
Twin-to-twin transfusion syndrome (TTTS) is a fetal malformation that affects 10–15% of monochorionic diamniotic pregnancies. The overall prevalence of TTTS is approximately 1 in 2000 pregnancies and usually occurs during the second trimester. TTTS has a variable and unpredictable course. If untreated, it is associated with a nearly 90% mortality rate for both fetuses.
Pathophysiology of TTTS
TTTS is caused by chorionic plate anastomoses between the two fetal circulations that cause unbalanced circulation exchange. Studies using a radiologic tracer have shown that inter-twin transfusion is nearly universal in TTTS. True connections between pairs of arteries (AA) or veins (VV) from the two fetal circulations are located on the chorionic plate. These anastomoses are bidirectional, and the net flow direction is determined by pressure differences between the circulations. Anastomoses between a chorionic vein and the twin’s chorionic artery lead to transfusion of blood from one twin to the other in a single direction and are referred to as arteriovenous (AV) anastomoses. These AV anastomoses are often multiple and balanced by other AV anastomoses in the opposite direction. TTTS is most often seen when AV anastomoses are present without AA anastomoses.
In TTTS, the donor twin becomes hypovolemic and oliguric, while the recipient twin becomes hypervolemic and polyuric. Because of these changes, the donor twin has activation of the renin-angiotensin system in an effort to preserve intravascular volume. This leads to hypertension, reduced placental perfusion, and growth retardation. On the other hand, the recipient twin has increased renal perfusion and urine output to counter the volume overload and also may be exposed to renin-angiotensin upregulation through placental shunts. The recipient commonly has cardiac abnormalities, including myocardial hypertrophy, increased velocities of pulmonic and aortic outflow, AV valve regurgitation, as well as right ventricular outflow obstruction and pulmonic stenosis, which may be from increased cardiac afterload caused by systemic hypertension.
Diagnosis of TTTS
Quintero staging for TTTS
Oligohydramnios in donor (DVP < 2 cm) and polyhydramnios in recipient (DVP > 8 cm)
Stage I plus no visible bladder in donor fetus
Stage II plus Doppler abnormality of reverse flow in the ductus venosus, absent or reverse end diastolic flow in the umbilical artery, or pulsatile flow in the umbilical vein
Stage II or III and hydrops fetalis in either fetus
Demise of one or both fetuses
Treatment of TTTS
The current mainstay for treatment of TTTS is fetoscopic selective laser photocoagulation (SLPC) targeting the anastomoses that contribute to the imbalance of flow, performed between 18 and 26 weeks gestation (Senat et al. 2004). Historically, amnioreduction was the primary treatment modality for TTTS but is now rarely applied as primary therapy unless TTTS develops outside the gestational age where SLPC is safe. The treatment for Stage I TTTS is a controversial subject because most Stage I patients do not progress to a later stage, but a trial of SLPC for Stage I disease is currently underway in Europe.
SLPC is performed percutaneously, most often under local anesthesia. A 2–3 mm fetoscope is inserted, with or without a trocar, under ultrasound guidance. The placental vasculature is mapped using direct visualization as well as Doppler ultrasound to identify anastomoses and to define the placental equator. All anastomoses between the two placental circulations are targeted for ablation with a 30–50 W diode laser. Amnioreduction may also be performed at the end of the procedure if necessary to reduce intrauterine pressure. Because of the incidence of both early and late complications of SLPC, close follow-up is important for all patients.
Outcomes of TTTS
The Eurofetus trial was a multicenter randomized controlled trial that compared serial amnioreduction to SLPC for TTTS. The laser therapy group had higher survival of at least one fetus to at least 28 days of age, 76% vs. 56% in the amniocentesis group. In addition, the laser group had a higher mean gestational age at delivery, with an average of 33 vs. 29 weeks in the amniocentesis group. Most importantly, at 6-month follow-up, the laser group had improved neurologic outcomes, with a decreased risk of periventricular leukomalacia.
While SLPC is much less invasive than it once was, there are still significant complications that accompany the procedure. Aside from the complications associated with fetoscopy itself, SLPC can be complicated by pseudoamniotic band sequence, TTTS recurrence, iatrogenic monoamnionicity, and twin anemia polycythemia sequence, which is defined as anemia in one fetus and polycythemia in the co-twin with normal AFV in both fetal sacs. A major long-term concern for TTTS survivors is neurodevelopmental abnormalities, affecting 6–25% of patients treated with SLPC (van Klink et al. 2016). These range from minor defects to major abnormalities including cerebral palsy, blindness, hemiparesis, and spastic quadriplegia.
Conclusion and Future Directions
Fetal surgery has seen dramatic progress in the last three decades, especially in the ability to diagnose, appropriately select, and treat fetuses with structural malformations that, if left untreated, would result in fetal demise or severely affect quality of life. In some cases, fetal surgery has clearly altered the natural history of the disease and improved outcomes, namely, CCAM, TTTS, and MMC.
In order for the field to continue to grow, several areas require continued study. First, maternal and fetal risk remain high, and renewed efforts to reduce morbidity and mortality associated with maternal-fetal intervention are paramount, including improvement in maternal tocolysis to control frequent preterm delivery. In addition, further innovations in endoscopic instrumentation and imaging modalities will contribute to more advanced minimally invasive approaches to replace open procedures. Furthering capabilities for image-guided interventions to safely permit diagnosis and treatment at even earlier gestational time points will decrease the risk of preterm labor and premature delivery. Randomized controlled trials when appropriate are essential to establish a clear benefit of maternal-fetal surgery for patients, allowing experimental therapies to move into clinical application.
- Danzer E, Hoffman C, D’Agostino JA, et al. Neurodevelopmental outcomes at 5 years of age in congenital diaphragmatic hernia. J Pediatr Surg. 2016; pii: S0022–3468(16)30284–6.Google Scholar