Prenatal regenerative fetoscopic interventions for congenital anomaliesBMJ 2020; 370 doi: https://doi.org/10.1136/bmj.m1624 (Published 01 July 2020) Cite this as: BMJ 2020;370:m1624
- Rodrigo Ruano, professor in obstetrics & gynecology, physiology and pediatric, and chair of Maternal-Fetal Medicine Division
- Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology and Center for Regenerative Medicine, Mayo Clinic College of Medicine, Rochester, Minnesota, USA
- Correspondence to R Rodrigo ;
Fetal intervention has progressed in the past two decades from experimental proof-of-concept to practice-adopted, life saving interventions in human fetuses with congenital anomalies. This progress is informed by advances in innovative research, prenatal diagnosis, and fetal surgical techniques. Invasive open hysterotomy, associated with notable maternal-fetal risks, is steadily replaced by less invasive fetoscopic alternatives. A better understanding of the natural history and pathophysiology of congenital diseases has advanced the prenatal regenerative paradigm. By altering the natural course of disease through regrowth or redevelopment of malformed fetal organs, prenatal regenerative medicine has transformed maternal-fetal care. This review discusses the uses of regenerative medicine in the prenatal diagnosis and management of three congenital diseases: congenital diaphragmatic hernia, lower urinary tract obstruction, and spina bifida.
Advances in early diagnostics have enabled therapeutic approaches before birth.1 An exciting output is the overarching goal of achieving regenerative correction prenatally, thereby averting florid disease before it becomes a chronic debilitating condition reliant on delayed postnatal clinical management. Indeed, congenital anomalies left untreated are associated with high mortality and morbidity. They are collectively responsible for a considerable number of perinatal deaths, and also carry serious emotional impact on families and economically on health systems. Infants with congenital anomalies need multidisciplinary care, including long duration management in tertiary centers and neonatal intensive care units. Of those who survive, most need complex postnatal surgeries and multidisciplinary medical care, especially when vital organs such as the heart, brain, lungs, and/or kidneys are affected.23
In utero interventions aim to avoid perinatal demise.1 The first described fetal therapy was intrauterine fetal blood transfusion for erythroblastosis fetalis, when antibodies from rhesus negative mothers cross the placenta and attack fetal rhesus positive red cells.4 These fetuses develop severe anemia, cardiac failure, and hydrops fetalis (anasarca) and then progress to in utero demise. Under ultrasound guidance, blood can be transfused to the umbilical cord to prevent cardiac failure and ensure fetal vitality.4
The potential to affect a congenital anomaly while the fetus is still linked to the placenta is now a reality. Clinical deployment of in utero interventions to address life threatening congenital diaphragmatic hernia (CDH), lower urinary tract obstruction (LUTO), and spina bifida are prime examples of how the field has evolved. Collectively, “prenatal regenerative therapies” have been proposed with the objective of reversing or preventing organ damage through restoration and/or regrowth of affected organs.5
This review summarizes evidence spanning more than 20 years that underpins the roll out of regenerative prenatal interventions for CDH, LUTO, and spina bifida. The growing experience with fetal therapy aims to restore fetal organ structure and function and improve postnatal outcomes, and provides the basis to expand the scope of regenerative medicine to a clinically applicable portfolio of prenatal interventions.
CDH Congenital diaphragmatic hernia
LUTO Lower urinary tract obstruction
FETO Fetal endoscopic tracheal occlusion
ELMO Extracorporeal membrane oxygenation therapy
LHR Lung to head ratio
MRI Magnetic resonance imaging
TFLV Total fetal lung volume
VAS Vesicoamniotic shunting
NTD Neural tube defect
MMC Meningomyelocele (open spins bifida)
Sources and selection criteria
We identified the references in this review through a comprehensive search of the following databases: Ovid MEDLINE(R) and Epub ahead of print, In-Process and other non-indexed citations, and Ovid EMBASE, Ovid Cochrane Central Register of Controlled Trials, and Scopus. An experienced librarian designed and conducted the search strategy with the input of the primary author. Controlled vocabulary was used to search for diagnostic methods, prognostic indicators, and fetal therapy for congenital diaphragmatic hernia, spina bifida, and congenital lower urinary tract obstruction. We included only English language articles, and we considered all study types—including case reports, case series, randomized controlled trials, reviews, and experimental animal studies—for this review. Articles outside of the topic of interest or beyond the scope of this review were excluded. The search strategies are available as appendix files (see supplementary files).
Guiding principles of fetal interventions
Fetal interventions were introduced to improve perinatal survival, but their role has expanded to improve morbidity, quality of life, and regenerate or restore malformed fetal organs.6 Proper selection of candidates is important to avoid unnecessary procedures in fetuses with highly fatal conditions or those with good prognosis without in utero intervention.78 The International Fetal Medicine and Surgery Society has proposed guiding principles including
- Precise prenatal diagnosis
- Well known pathophysiology of the congenital anomaly
- Absence of genetic/chromosomal disease
- Absence of associated major anomalies
- Natural course of the congenital anomaly showing a life threatening situation or even severe debilitation despite postnatal management
- Morbidity of the intervention should be acceptable for the mother and fetus
- Multidisciplinary evaluation and consensus
- A family informed consent after extensive counselling about potential risks and benefits of the proposed procedure
- Absence of adequate postnatal treatment for the condition
- Feasible in utero intervention, and
- Implementation of adequate ethical principles.9
To accomplish these principles requires rigorous experimental and clinical research studies.10 Experimental research and observational studies are important to understand the history and pathophysiology of congenital anomalies.11 Thereafter, investigations of possible in utero interventions are undertaken to establish feasibility, safety, and efficacy in clinical settings.12 Fetal surgeries have progressed from open hysterotomies to less invasive fetoscopic techniques with the objective of improving maternal and fetal outcomes.13
Fetal endoscopic surgeries
Fetal endoscopic surgeries, or fetoscopic procedures, are considered to be less invasive compared with open fetal surgery with hysterotomy. The aim is to improve survival and reduce morbidity by regenerating (restoring) malformed fetal organs relying on innate corrective healing or re-differentiating processes. This concept is illustrated in fetal endoscopic tracheal occlusion (FETO) for CDH, fetal intervention for lower urinary tract obstruction (LUTO); and fetoscopic spina bifida repair.
Fetal endoscopic tracheal occlusion for congenital diaphragmatic hernia
Definition and epidemiology of congenital diaphragmatic hernia
CDH is a failure of the diaphragm to fully close during early development, resulting in the herniation of abdominal contents into the chest cavity. CDH is associated with high mortality because of pulmonary hypoplasia and pulmonary hypertension (fig 1). The prevalence of CDH is approximately 2.5/10 000 pregnancies and isolated CDH has a predilection for male fetuses (1.5:1 male to female).1415 Defects are more common on the left side (80-90%) and occur in the posterolateral portion (90%).151617
CDH is often associated with major structural abnormalities and chromosomal aneuploidy (complex CDH). Major structural abnormalities are seen in 28-34% of fetuses, with cardiac, nervous system, and musculoskeletal being the most common14161718; however the reported incidence of complex CDH can be high as 55-61%.1920 Additionally, 8-18% of fetuses with CDH have aneuploidy, with trisomy 18 being most common, or other genetic syndromes such as Fryn’s or Apert syndrome.142021
Pathophysiology of congenital diaphragmatic hernia
Formation of the diaphragm occurs between the fourth and 10th week of gestation.22 The course of CDH occurs in the setting of aberrant formation of the pleuroperitoneal folds or the post-hepatic mesenchymal plate.2223 The predominant complications associated with CDH are pulmonary hypoplasia and pulmonary arterial hypertension.2425 Initially it was suspected that these complications occur as a result of compression from the abdominal contents protruding into the chest; however other data suggest a complex pathogenic process.
Lung development occurs in four stages starting at 4 weeks’ gestation and continuing into the postnatal period.2627 Lung budding is seen from 16 to 24 weeks’ gestation, and alveolar formation and maturation from 24 weeks’ gestation to 3 years of postnatal life.26 Concurrent with this process is pulmonary vascular development which occurs through neovascularization and branching of pre-existing conduits between weeks 10 and 11.27282930 Initially the pulmonary vessels are thick and muscular with high vascular resistance that decreases with gestation and after birth to promote gas exchange.27 In fetuses with CDH, on histologic exam the lungs appear less developed for gestational age, with the ipsilateral more severely affected than the contralateral lung.31 Decreased branching of bronchioles with subsequent acinar hypoplasia is apparent.313233 Fetal lungs affected by CDH from 17 to 24 weeks also show a decreased number of vessels and total vascular bed volume.3435 Additionally, pulmonary vessels are hyper-muscularized, with variable vascular reactivity contributing to postnatal pulmonary hypertension.3637
Advances in postnatal care of infants with CDH have reduced postnatal mortality. However, affected infants require complex postnatal treatment, including extracorporeal membrane oxygenation therapy (ECMO) and prolonged neonatal intensive care unit admission. Despite advances in postnatal management, postnatal mortality is 20-40%. Therefore, prenatal interventions for CDH were proposed to promote lung growth and development while the fetus is still oxygenated by the placenta.
Prenatal diagnosis and prognostication of fetuses with congenital diaphragmatic hernia
Outcomes in children with CDH are dependent upon the degree of pulmonary hypoplasia and pulmonary arterial hypertension. Previously, prenatal ultrasound was used to identify these lesions, which allowed for planning for delivery and immediate resuscitation efforts.3839 Prenatally diagnosed CDH tends to be more severe or complex (associated with other malformations) with lower long term survival than in infants diagnosed postnatally (73% versus 93%).40
While ongoing efforts to improve detection rates are needed, additional goals of prenatal evaluation have focused on prognostication markers for counseling prospective parents and identifying high risk fetuses that could benefit from early in utero intervention.4142 Different prognostic indicators have been proposed, but the most commonly used include the observed-to-expected lung-to-head ratio (o/e LHR) and observed-to-expected total lung volume (o/e TLV) as well as liver herniation (table 1).
Pulmonary hypoplasia and pulmonary arterial hypertension
Pulmonary hypoplasia and pulmonary arterial hypertension are the predominant complications associated with CDH morbidity and mortality. Many studies have evaluated how to best measure and predict lung volume. One study44 described the lung-to-head circumference ratio (LHR) which is obtained by tracing the contralateral lung area in a trans-axial view at the level of the four chamber heart and dividing it by the head circumference.44 It was subsequently noted that the lung grew more rapidly than the head circumference, leading to increasing LHR throughout gestation, however, the LHR in CDH fetuses compared with normal gestational age equivalents remains stable throughout pregnancy.4546 This o/e LHR has become the most utilized ultrasound tool to predict postnatal morbidity and mortality in clinical practice. Survival of neonates with isolated CDH based on o/e LHR of less than 25%, 25-45%, and greater than 45% was 18%, 66%, and 89%, respectively.4647 Other studies have confirmed a notable improvement in survival with increasing o/e LHR.43484950 Assessment of the liver position (being up or down), or liver-to-thorax ratio, in combination with o/e LHR improves predictability of postnatal outcomes.515253
Liver position (intra-abdominal or intrathoracic) also correlates with postnatal outcomes. Initial classification of liver-up or -down determined by color-flow Doppler ultrasonography was used to predict postnatal morbidity and survival.51545556 Additionally, studies have shown an association between the need for ECMO (25% versus 80%)55 and persistent pulmonary hypertension (12% versus 40%) in liver-down compared with liver-up, respectively.56 Magnetic resonance imaging (MRI) of the fetus allows for quantification of liver herniation, with higher herniated volume associated with mortality and ECMO requirement.5758
Postnatal identification of the stomach in the chest was a known predictor of increased mortality, which led to it being used as a prenatal marker. Subsequent studies have further refined prognostication by stomach location with a four-grade system, assessed in a trans-axial image at the level of the four chamber heart. This grading system in order of severity is described as: intra-abdominal, anterior left chest, mid-to-posterior chest, and retro-cardiac.5960 Postnatal morbidity and mortality correlate with stomach position grade.565961 The association with stomach position and mortality also appears to correlate with the degree of liver herniation.62
Total fetal lung volume
MRI of the fetus has enabled evaluation of total fetal lung volume (TFLV).63 Similar to LHR, TFLV is usually compared with the expected for gestational age (o/e TFLV). A meta-analysis showed correlation between o/e TFLV and postnatal mortality.64 Survival with o/e TFLV less than 35% ranges from 0% to 25% compared with 75-89% survival with o/e TFLV greater than 35%.49586566 In addition, one study58 showed improved prognostication when percentage liver herniation is incorporated with o/e TFLV.58 If MRI is not available, 3D ultrasound can also be used to calculate TFLV.6768
The above studies have evaluated markers that appear to correlate with the degree of pulmonary hypoplasia. One study69 used 3D power Doppler ultrasound to evaluate pulmonary vascular indices and showed a correlation between these indices and the severity of postnatal pulmonary arterial hypertension.69 This can serve as an additional tool to evaluate fetuses that may benefit from in utero intervention.
Fetal interventions for congenital diaphragmatic hernia
In the 1980s, it was identified that exposure of pregnant rodents to nitrofen (2, 4-dichloro-1(4-nitrophenoxy)benzene) resulted in varying degrees of CDH.7071 This enabled research into the course of CDH to better understand the embryology and pathophysiology of the disease. Lamb and rabbit models were introduced, in which an open in utero surgical procedure was performed to create a defect, which was subsequently closed and the pregnancy monitored for the remainder of gestation.7273 Current in utero treatment for CDH would not exist without these models.
The hypothesis that transitory tracheal fetal occlusion could prevent severe pulmonary hypoplasia was based on the clinical observation of congenital laryngeal/tracheal atresia. Fetuses with these conditions develop hyperplastic lungs, a finding that could address lung hypoplasia in fetuses with CDH.7475767778 Imposing fetal tracheal occlusion has been tested in the experimental lamb model including open neck dissection, external metal clips, or occlusion with silicone balloon followed by measuring fetal lung response.7679808182838485 These models proved that temporary occlusion trapped pulmonary secretions in the airways, which led to stretching of the pulmonary tissue and reversal of vascular changes.328687 With success in animal models, procedures were refined and translated into clinical practice.
Details of open in utero repair in animals and subsequently in humans have been documented.242588899091929394 These included testing the feasibility of closing the defect via an open approach. In those with liver-up, however, trying to replace the liver intra-abdominally led to obstruction of the ductus venosus blood flow and in utero demise.94 A prospective non-randomized trial88 that evaluated postnatal outcomes of open in utero repair of CDH found no benefit when compared with traditional postnatal repair; thus, the idea of open in utero repair was abandoned.88 With advancement of ultrasonography and fetoscopic instruments came the development of the FETO procedure. Compared with the open approach, this procedure could be done with a small port, which limited the risk of uterine rupture.808995 Initial surgeries9596 were performed with a low transverse skin incision and hysterotomy. The procedure was then refined9798 and a FETO technique developed that could be performed entirely percutaneously. The fetoscope is advanced into the uterus, at which point the surgeon directs the scope into the fetal mouth, larynx, and then trachea. When in the appropriate position, a detachable balloon is inflated and deployed (fig 2A). The balloon is left in place for several weeks of pregnancy. Removal was initially performed using the ex utero intrapartum treatment procedure (EXIT).9699 The current standard for balloon retrieval includes fetoscopic balloon removal at around 34 weeks, percutaneous ultrasound guided puncture, or postnatal tracheoscopy immediately after birth.959798100101102103 Prenatal or immediate postnatal deflation of the tracheal balloon is essential to avoid neonatal demise.
The initial FETO randomized control trial was performed in fetuses with moderate to severe disease (defined as an LHR less than 1.4 measured between 22 and 28 weeks with liver herniation) through laparotomy and hysterotomy.96 This trial found a higher frequency of prematurity (73% in the fetal tracheal occlusion group versus 31% in the standard postnatal treatment group, P=0.10) and no significant benefit in the FETO group (P=1.00), possibly related to hysterotomy; however the standard of care group had a higher than expected survival rate96 (survival rate of 73% in the fetal tracheal occlusion group versus 77% in the standard postnatal treatment group, P=1.00). Subsequently, the technique was refined97 to a percutaneous method, and showed statistically significant improvement in perinatal morbidity and mortality with FETO. A prospective non-randomized study98 of 210 fetuses who underwent FETO confirmed the feasibility of the procedure with successful balloon placement on first attempt in 97% of cases, and median procedure duration of 10 minutes. The study found improved survival compared with what was expected based on disease severity. The survival increased significantly in concordance with higher o/e LHR (survival rate (%)=(258×(o/e LHR (%))−28.68)/100; r=0.974, P<0.0001.98
A randomized trial104 included 41 fetuses with severe CDH defined as LHR <1.0 with liver herniation. It found that survival at 6 months was significantly increased from 5% to 50% in the standard FETO group for severe CDH (with balloon placement between 26 and 29 weeks’ gestation) (relative risk10.5 (95% confidence interval 1.5 to 74.7) P<0.01).104 Another trial101 also showed the feasibility of early FETO and improved outcomes in fetuses with extremely severe CDH (defined as LHR <0.7 or o/e LHR <0.17) with balloon placement as early as 22-24 weeks, compared with the standard balloon placement at 26-30 weeks. Survival of infants with extremely severe CDH with early placement was 63%, compared with 11% for the standard timing, and 0% for fetuses without FETO (P<0.01).101
Some studies have even suggested that FETO may improve fetal pulmonary vasculature and therefore it prevents severe pulmonary arterial hypertension.100105 A large European multicenter randomized controlled trial (Tracheal Occlusion To Accelerate Lung Growth, TOTAL), is recruiting patients to confirm the benefits of FETO for fetuses with severe left sided CDH (ClinicalTrials.gov Identifier: NCT01240057) as well as moderate CDH (ClinicalTrials.gov Identifier: NCT02875860).
The most common complications associated with FETO are premature pre-labor rupture of membranes and preterm delivery, ranging from 36% to 47% and 31% to 42%, respectively.9798101104 Ongoing trials will continue to provide information to optimize timing and safety of the FETO procedure.
Prenatal regenerative therapy for congenital diaphragmatic hernia
We have proposed that FETO serve as a prenatal regenerative therapy for CDH by promoting fetal lung regrowth and organ redevelopment.5 Our studies have shown that FETO performed at 28 weeks for severe CDH (o/e LHR between 20% and 25%) promotes fetal lung growth until a plateau of maximal growth six weeks after the procedure (at 34 weeks)100 (fig 2B). Our studies have also suggested that FETO improves pulmonary vasculature status associated with decreased risk of pulmonary arterial hypertension.100103 Adequate fetal lung growth and pulmonary response were associated with increased survival rate. However, minimal fetal lung response was observed in fetuses with extremely severe CDH who underwent FETO at 28 weeks.100
A multicenter clinical trial compared early FETO (at 22-24 weeks) with classic FETO (at 26-30 weeks) in 27 fetuses with extremely severe CDH (o/e LHR <17%),101 showing that early FETO was associated with a higher survival rate as a consequence of a better fetal pulmonary response (survival rate of 62.5% in the early FETO group, 11.1% in the classic FETO group, and 0% in the postnatal standard treatment group, P<0.01).101 The response was not only more pronounced, but also persisted longer, with a plateau of maximal fetal pulmonary response achieved at 8-10 weeks after the procedure (around 32 weeks).101 This finding in early FETO can be explained by the physiology and embryological development of the lung in utero.101 Early FETO is performed during the end of the canalicular period of lung development, during which the terminal bronchioles, respiratory bronchioles, alveolar ducts, and capillary network are formed. Based on our experience, we propose early FETO for extremely severe CDH (now at o/e LHR <20%), and classical FETO for severe CDH (o/e LHR between 20% and 25%). Currently, our group does not offer FETO for mild CDH (table 1).
Based on the fetal pulmonary response, FETO may promote fetal lung regeneration.103 Animal and clinical studies have confirmed that FETO promotes fetal lung hyperplasia and growth; however, the exact mechanisms involved in fetal lung regeneration after FETO remains unknown. Future studies are necessary to investigate fetal lung development after FETO.106107108109
Fetal cystoscopy for lower urinary tract obstruction
Definition and epidemiology of lower urinary tract obstruction
Fetal LUTO is defined as bladder outlet obstruction resulting from congenital renal outflow tract anomalies, and affects two to three infants per 10 0000 live births. The prevalence of antenatal LUTO is likely much higher, however, when considering elective termination and intrauterine fetal death.
LUTO represents a heterogeneous group of urinary outflow tract anomalies, each with distinct prevalence and/or gender predilection. It can occur as an isolated defect (isolated LUTO) or may be accompanied by other congenital abnormalities (complex LUTO) including chromosomal, cardiac, rectal, brain/spine, or skeletal anomalies.110111 Posterior urethral valve is the most common cause of LUTO (63%). Less common causes include urethral atresia, urethral stenosis, prune belly syndrome, and “unspecified” LUTO (9.9%, 7%, 2.5%, and 17.6%, respectively).110112 Previous studies have also shown an association between gestational age at diagnosis and a high prevalence of less common LUTO causes. such as urethral atresia.113 Etiological classification of LUTO not only provides specific prevalence and outcome data, but is also of prognostic significance correlating with survival. In addition, the wide range of manifestations seen in LUTO likely represent a variation in severity of outflow tract obstruction (complete versus partial bladder outlet obstruction).114
The high perinatal morbidity and mortality of LUTO are primarily owing to complications of renal impairment, oligohydramnios, and pulmonary hypoplasia.111114 LUTO accounts for 15% to 20% of pediatric end stage renal failure115 as well as 10% to 60% of pediatric renal transplantations.111 Mortality rates vary but can be as high as 80% to 90%.114116
Pathophysiology of lower urinary tract obstruction
LUTO results in fetal bladder dilation (megacystis) with subsequent bladder muscle hypertrophy/hyperplasia and increased intravesical pressure and hydro-ureter and hydronephrosis.117 Urinary stasis in the renal pelvis and calyces leads to renal dysplasia.117 The downstream effects of fetal anuria and in utero renal dysfunction include oligohydramnios (or anhydramnios) and pulmonary hypoplasia.117118 Animal models of LUTO have documented the natural course of LUTO in fetal rabbit, fetal rat, pigs, and guinea pigs. The fetal lamb model has been the most reliable.119 These studies119120 have confirmed the downstream effects of bladder outflow obstruction. According to one study,121 histologic analysis of the obstructed kidneys in LUTO shows cystic dysplasia in the sub-capsular renal cortex, dilated primitive ductules with fibrous tissue cuffs, primitive glomeruli, and disorganized interstitia. The pattern of renal dysplasia documented in these animal studies is similar to that of human neonates,120122123 which suggests a possible causal link between LUTO and renal dysplasia.
The rationale for fetal intervention in LUTO is based on the understanding of the natural course and detrimental outcomes of LUTO. Fetal therapy has the potential to ameliorate pulmonary hypoplasia and possibly prevent end stage renal disease.
Prenatal diagnosis and prognostication of fetuses with lower urinary tract obstruction
Prenatal detection and prognostication of LUTO facilitates proper counselling of potential parents as well as appropriate selection of candidates for fetal therapy. Fetal anatomic ultrasound screening has improved prenatal detection of LUTO. According to a large population study, prenatal detection of LUTO increased from 33% to 62% over the 14 year study period.111 Comprehensive fetal anatomic survey is also warranted to rule out complex LUTO, which makes up approximately 20% of all cases.110111 Hence, fetal echocardiography, genetic counselling, chorionic villus sampling, or amniocentesis are all essential components of investigating LUTO depending on initial presentation or clinical suspicion.
The sensitivity of ultrasound diagnosis of LUTO is between 50% and 59%, according to two large LUTO population series110111 but sensitivity as high as 95% has been reported for certain ultrasound parameters, such as renal hyper-echogenicity.124 Ultrasound findings in LUTO include oligohydramnios (or anhydramnios), which is defined as an amniotic fluid index <5 cm or maximum vertical pocket <2 cm,125 dilated bladder with thickened wall, “key hole” sign, ureteral dilation, hydronephrosis, renal hyper-echogenicity, subcortical renal cysts, or renal dysplasia124 (fig 3A).
Dilated bladder (sensitivity, 96.8% P<0.001) and thickened bladder wall (sensitivity, 93.5%, P<0.001) were the best sonographic indicators of LUTO, according to a retrospective cohort study.126 Fetal ultrasound can also be used to prognosticate the outcomes of fetuses with LUTO. A systematic review of the accuracy of prenatal ultrasound in fetuses with LUTO showed a high sensitivity and specificity of the following sonographic parameters in predicting outcome: oligohydramnios, renal cortical appearance, and early diagnosis of LUTO (before 24 weeks’ gestation). Renal cortical appearance was the most predictive of postnatal renal function (sensitivity 0.57 (95% confidence interval 0.37 to 0.76); specificity 0.84 (95% confidence interval 0.71 to 0.94); area under curve 0.78).127 Prenatal ultrasound also helps to rule out other fetal anomalies. However, definitive diagnosis or prognosis of LUTO cannot be provided by ultrasound alone.114124
Fetal urine biochemistry is another component of LUTO investigation that can be used to estimate fetal renal function. It involves fetal urine sampling, urine analysis, and interpretation of urinary electrolyte levels. According to a prospective cohort study of 24 LUTO patients (and 26 controls), urinary electrolytes decreased and urinary creatinine increased with gestational age in normal fetus controls (likely because of normal fetal renal system maturation).128 Conversely, LUTO patients with or without renal dysplasia had statistically significantly higher levels of electrolytes when compared with controls. Furthermore, LUTO fetuses with renal dysplasia had higher levels of urinary electrolytes and β2-microglobulin.128 A reference range for urine biochemistry was suggested as “favorable” and hence amenable to fetal surgery when urine osmolarity is <200 mOsm/L, sodium <100 mEq/L, chloride <90 mEq/L, and β2-microglobulin <6 mg/L.114129130 Reports in the literature are conflicting, however. One meta-analysis showed low sensitivity and specificity of fetal urinary biochemistry in predicting postnatal renal function.131 Conversely, a retrospective study of 72 fetuses with megacystis showed a statistically significant correlation between fetal urinary electrolytes and postnatal renal outcomes.132 The use of fetal urinary biochemistry to evaluate fetal renal function should, therefore, take into account three main considerations: (1) the available evidence is mostly based on studies that evaluated fetal urinary biochemistry during the second trimester; (2) fetal urinary biochemistry is only a “snapshot picture” of fetal renal function, which means that many other factors could influence renal function prenatally and postnatally; and (3) fetal urinary biochemistry is only a part of comprehensive fetal evaluation and should be interpreted along with fetal renal imaging for appropriate staging of disease severity.112133134135136137
Accurate prenatal detection and prognostication of LUTO is vital to the appropriate selection of fetal intervention candidates to optimize perinatal outcomes. To this end, standardized prenatal LUTO diagnosis using prenatal ultrasound, fetal urine biochemistry, and clinical parameters has been proposed to improve LUTO prognostication.138139140141 A retrospective study133134138 described a prenatal LUTO staging system that correlates with postnatal survival and also proposed fetal intervention according to disease severity (table 2).
Similar findings were confirmed by a large retrospective study of 261 LUTO patients who were managed conservatively.139 A clinical scoring system that combines specific sonographic findings, clinical parameters, and fetal urinary biochemistry is under investigation.142
Fetal interventions for lower urinary tract obstruction
The main aims of fetal intervention are to prevent severe pulmonary hypoplasia and end stage renal disease (for stage II LUTO). Fetal urine can be collected by fetal vesicocentesis under ultrasound guidance. Sequential urine sampling (up to three samples) over a 24-48 hour interval (to avoid repeat sampling of stagnant urine) is preferred to better reflect renal function.114131138 Following proper risk stratification and disease prognostication, fetal intervention may be considered depending on LUTO severity (table 2).
Vesicoamniotic shunting (VAS) aims for sustained bladder decompression with the help of a bladder catheter which allows continuous bladder drainage for the remainder of gestation. The Percutaneous Vesicoamniotic Shunting versus Conservative Management for Fetal Lower Urinary Tract Obstruction (PLUTO) randomized controlled trial assessed the effectiveness of VAS for treating LUTO.143 This trial compared 16 patients who underwent VAS with 15 patients managed conservatively and showed significantly higher neonatal survival to 28 days of life with VAS (relative risk 3.2, 95% CI 1.06 to 9.62; P=0.03). However, the study failed to show a long term reduction in pediatric morbidity at 1 to 2 years of age, and the trial was halted early because of poor recruitment. Robust standardization of LUTO investigation could potentially address the problem of poor recruitment and help stratify management in order to accurately assess the outcomes of fetal intervention in future prospective LUTO trials.138139
A meta-analysis including 112 LUTO patients treated with VAS (and 134 controls), found that VAS was associated with higher perinatal survival (57.1% with VAS versus 38.8% with conservative management, P<0.01). However, there were no significant differences in 6-12 month survival, two year-survival, or postnatal renal function in treatment compared with control groups.144 Our protocol is investigating the benefits of fetal VAS for Stages II (to prevent severe pulmonary hypoplasia and end stage renal disease) and III LUTO (to prevent severe pulmonary hypoplasia as a bridge to renal dialysis/transplantation).
Fetal cystoscopy involves direct visualization of the urinary outflow tract (with a fetoscope) for etiological diagnosis and specific treatment of LUTO113133134135136137144145146147148 (fig 3B). The fetoscope is placed into the bladder outlet (under ultrasound guidance) and pediatric posterior urethral valve (PUV), the commonest etiology of LUTO, can be treated with laser fulguration of occluding membranes.114
Fetal cystoscopy is still under investigation, but initial results are optimistic. According to a systematic review of non-randomized trials that evaluated fetal cystoscopy as a diagnostic and therapeutic modality for LUTO, the sensitivity was 100% to correctly diagnose the cause of LUTO. Compared with conservative management, fetal cystoscopy was associated with higher perinatal survival (odds ratio 20.51 (95% CI, 3.87 to 108.69)).145 A multicenter study comparing VAS, cystoscopy, and conservative management revealed that although VAS and fetal cystoscopy improved six month survival rate, only fetal cystoscopy significantly improved postnatal renal function (absolute risk reduction 2.66 (95% CI, 1.25 to 5.70)),112 probably related to better patient selection. However, fetal cystoscopy is a challenging procedure associated with some technical limitations that can lead to complications such as urological fistulas (in approximately 10% of the cases) and prematurity with a mean gestational age at delivery of 34.6 ± 2.5 weeks (range 28–37 weeks).137147 Urological fistulas are usually caused by inadequate curvature of the instrument, fetal mobilization, and limited surgeon experience.147 The long term benefits of fetal cystoscopy over VAS remain undetermined and improvements of the instruments and techniques are warranted.149 Our group is currently investigating the benefits, safety, and risks related to fetal cystoscopy for stage II LUTO (ClinicalTrials.gov Identifier: NCT03281798).
Serial amnioinfusion, which involves repeated infusion of sterile warm saline or lactated Ringer’s solution to restore amniotic fluid, is indicated for stage IV LUTO (intrauterine fetal renal failure) that occurs spontaneously or after fetal VAS (in stage III LUTO), according to our protocol (ClinicalTrials.gov Identifiers: NCT03723564, NCT03101891133134138). It aims to prevent severe pulmonary hypoplasia and perinatal demise, working as a bridge to postnatal dialysis and renal transplantation.150151 However, ethical and clinical questions persist on its benefits and safety particularly because renal transplantation cannot usually be offered to children under 2.152 For this reason, serial amnioinfusion is still under investigation,151 and clinical trials are under way (ClinicalTrials.gov Identifiers: NCT03723564, NCT03101891).
Some investigators have suggested treating stage I LUTO (with normal amount of amniotic fluid).153154 However, this suggestion is controversial because these newborns usually don’t have severe pulmonary hypoplasia or progress to end stage renal disease. No evidence suggests that fetal interventions can prevent renal damage in these infants; fetal interventions are associated with obstetrical complications.
Prenatal regenerative therapy for lower urinary tract obstruction
We have proposed a new concept of prenatal regenerative therapy for LUTO.5 “Regenerative prophylaxis” in LUTO involves possible restoration of fetal renal function preserving organ development. The ability of VAS and fetal cystoscopy to decompress the fetal urinary tract in LUTO, restore amniotic fluid volume, and therefore possibly promote better fetal lung and renal development may be considered as a form of regenerative medicine.5 Along with serial amnioinfusions, these therapies provide pulmonary palliation and may promote adequate lung development through unknown mechanisms.
Fetal endoscopic repair of spina bifida
Definition and epidemiology of spina bifida
Neural tube defect (NTD) describes congenital malformations of the central nervous system (CNS) that occur secondary to lack of the neural tube closure during early development. Spina bifida is the most common non-lethal congenital defect of the central nervous system.155
The incidence of neural tube defects ranges from 1.0 to 10.0 per 1000 births. The estimated birth prevalence of spina bifida in the US is 3.5 per 10 000 live births. The Centers for Disease Control and Prevention (CDC) report that Hispanic people have the highest rate (3.80 per 10 000 live births), when compared with non-Hispanic white people (3.09 per 10 000 live births) and non-Hispanic black people (2.73 per 10 000 live births).156157
Epidemiological studies discovered that maternal folate status is critical for proper neural tube closure during embryogenesis.158 This prompted the US Public Health Service to recommend 400 µg of folic daily acid in women considering pregnancy to prevent NTDs. This has resulted in a statistically significant reduction in the prevalence of NTDs.159 Despite this recommendation and advancements in diagnosis and postnatal management, however, spina bifida remains a major source of morbidity and mortality.160
Pathophysiology of spina bifida
Open spina bifida or meningomyelocele (MMC) is characterized by failure of the neural tube closure with herniation of the meninges and spinal cord through a vertebral arch defect. This results in lifelong motor, sensory, and neurodevelopmental disabilities. The severity and extent of the disease is defined by the upper level of the anatomic defect161 and can range from bladder, bowel, and sexual dysfunction, to involvement of the lower and even upper extremities with secondary orthopedic disabilities.160
The pathophysiology is characterized by a “two hit” process, which is initiated by the failure of the posterior neuropore closure, followed by inflammatory and traumatic spinal cord damage resulting from amniotic fluid toxicity in utero. Children affected by this condition also invariably have an associated Arnold Chiari II malformation (or hindbrain herniation) possibly as a result of cerebrospinal fluid leakage, leading to progressive downward displacement of the hindbrain. This malformation is also associated with hydrocephalus and developmental brain abnormalities162163 (fig 4A). The rationale for fetal intervention is to prevent the “second hit” and therefore limit inflammation and downstream effects of MMC.
According to data from the US Spina Bifida Registry, approximately 80% of MMC patients underwent ventriculo-peritoneal shunt placement to treat hydrocephalus, 96% had impaired bladder function, 92% had bowel dysfunction, and close to 40% were not able to walk.164 Data from the Danish database showed that 7% of MMC patients died in the first year of life secondary to pneumonia, meningitis, peritonitis, pyelonephritis, or sepsis.165 These data highlight the significant morbidity and mortality associated with MMC.
Prenatal diagnosis and prognostication of fetuses with spina bifida
MMC is diagnosed on routine second trimester ultrasound. It is often identified in the sagittal plain as a cystic lesion on the posterior spine with varying degrees of lumbosacral vertebral distortion.166 Associated cranial features, including ventriculomegaly, microcephaly, frontal bone scalloping (“lemon” sign), an abnormal posterior curvature of the cerebellum (“banana” sign), or “absent” cerebellum, may also be seen on ultrasound.166167 Given the correlation between MMC level and disease severity, functional sonographic evaluation of lesion level has been proposed161 as a predictor of postnatal ambulation prognosis. In a prospective study, the authors evaluated the segmental lesion level based on the most distal active muscle movement on antenatal ultrasound (table 3) and postnatal evaluation. The agreement between the designated prenatal and postnatal segmental levels was 91.7% and 88.9% for the right and left limbs, respectively.161
Fetal MRI is another important component of a comprehensive investigation of spina bifida. It provides a detailed assessment of the fetal spine and brain and also helps to rule out other associated anomalies. The presence or absence of an MMC sac, the size of the sac (when present), and size of the vertebral arch defect are all relevant spine MRI parameters; the presence or absence of hindbrain herniation, degree of hindbrain herniation (when present), and ventricular size are important brain MRI parameters.168 These findings are all of prognostic significance as they guide the approach to management. Indeed, the inclusion or exclusion criteria for prenatal treatment of MMC, adopted by the Management of Myelomeningocele Study (MOMS) trial and other fetal centers worldwide, are based on prenatal diagnostic parameters.
Fetal interventions for spina bifida
Historically, MMC was repaired postnatally with surgical closure of the lesion and ventricular shunt placement for hydrocephalus treatment. Recently, studies have shown benefits of in utero intervention for MMC.163169 Prenatal MMC repair was first performed in humans in 1997.
Early data suggested a dramatic improvement in hindbrain herniation of the fetuses, but with an inherent risk of preterm birth, uterine dehiscence, fetal or neonatal death. Further investigation with the MOMS trial,163 which compared in utero MMC closure with routine postnatal repair, showed that prenatal repair significantly decreased need for shunting, reversed hindbrain herniation, and improved neurologic function when compared with postnatal repair.169 The trial highlighted several benefits of prenatal repair, including a 50% reduction in the need for postnatal shunt placement (P<0.001); at 12 months of age, 36% of infants in the prenatal repair group had no hindbrain herniation compared with only 4% in the postnatal repair group. Children in the prenatal surgery group were also more likely to walk without orthotics (42% versus 21%) and had better motor function. Importantly, those in the prenatal repair group had a higher Bayley Psychomotor Development Index score.170171
However, the MOMS trial also reported several important complications associated with in-utero open repair of MMC. There was a significantly increased risk of preterm delivery, premature rupture of membranes. and uterine dehiscence (only 64% had an intact well healed hysterotomy site at the time of planned caesarean section). Maternal risks included approximately 6% risk of pulmonary edema (attributed to use of tocolytics), 9% risk of blood transfusion at delivery, and need for a caesarean section for all future pregnancies, irrespective of other obstetric indications.163 Other studies have also found evidence of myometrium scarring and substantial thinning or dehiscence of the hysterotomy after open MMC repair.172 Tocolytics have been used preoperatively and postoperatively to minimize the risk of preterm birth after open MMC repairs.173
Recognizing the conflicting outcomes of neonatal benefit versus maternal morbidity has led to the exploration of fetoscopy as a less invasive approach to in utero MMC repair. Several attempts have been made to improve techniques and clinical outcomes for fetoscopic repair, and some groups perform this procedure completely percutaneously while others opt for maternal laparotomy with fetoscopy. Fetoscopic repair of MMC is still under investigation to evaluate benefits, safety, and technical aspects.174175
The first described literature on in utero fetoscopic repair of MMC in humans dates back to 1997.176 Here, the authors described two cases where a maternal split-thickness skin graft was placed over the exposed neural placode. One fetus died from complications of prematurity and the other survived following a planned caesarean delivery at 35 weeks’ gestation. The same authors performed a non-randomized trial study in 2000 comparing outcomes of fetoscopic versus open repair of MMC with four fetuses in each group. They found that the open repair group delivered at a later gestational age had a shorter operative time and better wound healing when compared with the fetoscopic repair group.177
Others175 have suggested draining some amniotic fluid and partially filling the uterus with carbon dioxide gas for better visualization before the surgical repair, using a completely percutaneous approach (fig 4B). This led to a concern for fetal acidemia and placental dysfunction from carbon dioxide exposure, as seen in sheep studies.178179180 A cohort study of patients who underwent fetoscopic repair of MMC181 evaluated venous cord blood gas in three fetuses before and after laparoscopic-fetoscopic repair with carbon dioxide insufflation and found that the partial pressure of oxygen and carbon dioxide remained in the normal range, suggesting that carbon dioxide insufflation during fetoscopic MMC repair does not cause acidemia in human fetuses. Another retrospective cohort study 182 looked at fetal growth outcomes following laparoscopic-fetoscopic MMC repair carbon dioxide insufflation versus open MMC repair. It found that infants exposed to fetoscopic or open MMC repair in utero did not show statistically significant differences in fetal or postnatal growth parameters.
Appropriate anesthesia protocol is paramount to the success of open and fetoscopic repair of MMC in utero. Data from a retrospective cohort study comparing anesthesia protocol in open with fetoscopic MMC repair found that open surgery was associated with higher doses of halogenated anesthetic agents, sevoflurane, increased need for intra-operative tocolytic drugs with nitroglycerine, and postoperative tocolysis with magnesium sulfate, and a higher volume of colloids.183 From a hemodynamic standpoint, median mean arterial pressure was lower in open versus fetoscopic surgery; systolic blood pressure, diastolic blood pressure, and mean blood pressure decreased during uterine exposure, and this descent was more acute in open surgery.183 These results suggest a possible advantage of fetoscopic over open MMC repair.
As techniques for fetoscopic repair have gradually improved over the years, so have the outcomes. While open in utero spina bifida has remained the standard approach, fetoscopic repair holds promising results for optimizing maternal obstetric outcomes, with the hope of maintaining similar fetal and neonatal outcomes.170184185 Future randomized controlled trials are necessary to confirm recent reports.
Prenatal regenerative therapy for spina bifida
Regenerative prophylaxis with respect to spina bifida involves the restoration of hindbrain anatomy in utero. Hindbrain herniation in MMC is a result of “cranio-spinal dissociation” as the normal interaction between cerebrospinal fluid (CSF) spaces of the cranium and spine is disrupted by CSF flow abnormalities.5155 Early MMC closure restores hindbrain herniation, which has been associated with a lower risk of hydrocephalus when compared with postnatal MMC closure.163 Small series have shown that in utero MMC closure improves hindbrain herniation prenatally, as early as 4-6 weeks postoperatively.168186
In the past 20 years, progress in fetal surgeries has been extraordinary, with refined techniques, indications, implementations, and applications. In utero procedures have increasingly become a part of perinatal options in tertiary centers specialized in the treatment of congenital anomalies. Our group has introduced the concept of fetal regenerative therapy, where fetal surgeries are implemented to promote restoration, growth, and regeneration of abnormally developed fetal organs, aimed at improving perinatal survival and reducing morbidity. The future of this field is promising, pending the results of ongoing clinical trials. Understanding the mechanisms involved in restoration or regeneration of fetal organs will also open an opportunity for even less invasive novel fetal regenerative therapies. A collaborative effort among medical specialties is necessary to foster the success of this evolving practice.
Can fetal surgeries restore fetal organ structures and functions as well as improve postnatal outcomes using less invasive techniques?
How patients were involved in the creation of this article
No patients were involved
Series explanation: State of the Art Reviews are commissioned on the basis of their relevance to academics and specialists in the US and internationally. For this reason they are written predominantly by US authors
Financial support and competing interests: I acknowledge the financial support from the State of Minnesota (RMM 102516008). Dr R. Ruano is a recipient of the Regenerative Medicine Minnesota Clinical Trial grant: “Fetoscopic Regenerative Therapy for Severe Pulmonary Hypoplasia – a feasibility pre-randomized control trial study.”
Acknowledgments: I thank Dr Eniola R. Ibirogba, Dr Kavita Narang, Dr Michelle Wyatt, and Dr Andre Terzic for their contributions to the content and review of this manuscript. I also thank Ms. Jan H. Case for her work in preparing illustrations and the Mayo Clinic Library staff for their support with the literature search.
Provenance and peer review: commissioned; externally peer reviewed.