Chapter 24Invasive Fetal Therapy Jan A. Deprest, MD, PhD, Eduardo Gratacos, MD, PhD, and Liesbeth Lewi, MD, PhDThe availability of high-resolution ultrasound imaging and screening cedure done by Quintero a few months earlier. The Leuven groupprograms has made the unborn child a true patient. When fetal mal- subsequently set up the Eurofoetus consortium supported by theformations, genetic diseases, or in utero acquired conditions are sus- European Commission (E.C.), acting as liaison between selectedpected, patients are referred to tertiary care units with more specialized European fetal medicine units and an endoscopic instrument makerskills, technical equipment, experience, and multidisciplinary coun- to design new endoscopes and instruments to improve managementselors to deﬁne potential options. In some cases, intervention before of speciﬁc conditions.4,5 The later execution of a successful randomizedbirth may be desirable, which often does not require direct access to trial on laser coagulation of twin-twin transfusion syndrome (TTTS)the fetus—for example, transplacental administration of pharmaco- prompted in Europe wide clinical acceptance of fetoscopy.logic agents for cardiac arrhythmias or antibiotics in case of fetal Progress in the practice of fetal surgery has been slowed by theinfection. Other conditions can be treated only by invasive access to paucity of randomized trials, by patient and practitioner reluctance,fetus. In utero transfusion of a hydropic fetus to treat the anemia of by availability of local operator expertise, and even by the lack of regu-Rh isoimmunization, ﬁrst described in 1961, was probably the ﬁrst latory approval of novel surgical instrumentation. For example, opensuccessful invasive therapeutic procedure. Today, blood transfusion fetal surgery can now be performed with signiﬁcantly improved out-through the umbilical cord, intrahepatic vein, or (exceptionally) comes and fewer side effects compared with a decade ago. Despite that,directly into the fetal heart is widely offered, with good fetal and such procedures are rarely done in Europe, with the exception oflong-term outcome when procedures are done by experienced operations on placental support.6 In the United States, slow regulatoryoperators. acceptance of new fetoscopic instruments has contributed to the con- Some conditions are amenable to surgical correction, and in the troversy regarding the place of fetoscopic treatment of TTTS,7,8 togethermajority of cases this is best done after birth. Occasionally, prenatal with a historically stronger preference for open procedures, includingsurgery is required to save the life of the fetus, or to prevent permanent even for nonethical conditions. Indeed, in utero repair of myelome-organ damage. This can be achieved by correcting the malformation, ningocele (MMC) is currently being performed in the MOMS trial (seeby arresting the progression of the disease, or by treating some of the Outcome of Antenatal Neural Tube Defect Repair, later), sponsoredimmediately life-threatening effects of the condition, delaying more by the National Institutes of Health, which tests the hypothesis thatdeﬁnitive repair until after birth. Because of the potential complica- antenatal repair will reduce morbidity in survivors compared withtions, risks and beneﬁts of the intervention must be weighed against postnatal repair. Should the results of the MOMS trial be positive,each other. Table 24-1 summarizes the indications and rationale for undoubtedly European centers will have to reconsider the issue ofin utero surgery on the fetus, placenta, cord, or membranes. open surgery, just as increasing experience and more encouraging A consensus, endorsed by the International Fetal Medicine and long-term results of fetoscopic procedures from Europe have spurredSurgery Society (IFMSS), has been reached on the criteria and indica- the spread of minimally invasive procedures throughout the Unitedtions for fetal surgery (Table 24-2).1 In the 1980s to 1990s, only a few States.conditions met these criteria, and surgical intervention requiredmaternal laparotomy, partial exteriorization of the fetus through astapled hysterotomy. These “open” procedures were initially associatedwith high fetal-maternal morbidity, raising the question for some of Open Fetal Surgerythe value of claimed beneﬁts. Open fetal surgery is a complex enterprise that should be undertaken The growing availability of videoendoscopic surgery in the 1990s, only in centers staffed with skilled personnel. Because of the high inci-combined with earlier experience with fetoscopy, paved the way for dence of preterm labor, prophylactic tocolysis is essential, using forthe concept of endoscopic fetal surgery. The rationale was that mini- instance indomethacin or nifedipine. Large-bore venous access ismally invasive access to the amniotic cavity would reduce the fre- established, but ﬂuid administration is conservative and meticulouslyquency of preterm labor and diminish maternal morbidity. Investigators managed to reduce the risk for pulmonary edema that frequently occursat the Centre for Surgical Technologies (CST) in Leuven, Belgium, with certain tocolytics. Open surgical procedures are typically per-have helped advance the application of these techniques by ﬁrst estab- formed using general endotracheal anesthesia, taking advantage of thelishing an ovine model for endoscopic fetal surgery.2 That experience myorelaxant and uterine contraction suppression qualities of haloge-laid the basis for the ﬁrst successful umbilical cord ligation in Europe,3 nated anesthetic gases. The uterus is exposed by a large laparotomy andalmost simultaneously with, but independently from, a successful pro- opened with specially designed, resorbable lactomer surgical staples
434 CHAPTER 24 Invasive Fetal TherapyTABLE 24-1 INDICATIONS AND RATIONALE FOR IN UTERO SURGERY ON THE FETUS, PLACENTA, CORD, OR MEMBRANES Pathophysiology Rationale for In Utero InterventionSurgery on the Fetus1. Congenital diaphragmatic hernia Pulmonary hypoplasia and anatomic substrate Reverse pulmonary hypoplasia and reduce degree of for pulmonary hypertension pulmonary hypertension; repair of actual defect delayed until after birth2. Lower urinary tract obstruction Progressive renal damage by obstructive Prevention of renal failure and pulmonary hypoplasia uropathy by anatomic correction or urinary deviation Pulmonary hypoplasia by oligohydramnios3. Sacrococcygeal teratoma High-output cardiac failure by arteriovenous Reduction of functional impact of the tumor by its shunting and/or bleeding ablation or (part of) its vasculature Direct anatomic effects of the tumoral mass Reduction of anatomic effects by draining cysts or bladder Polyhydramnios-related preterm labor Amnioreduction preventing obstetric complications4. Thoracic space-occupying lesions Pulmonary hypoplasia (space-occupying mass) Creating space for lung development Hydrops by impaired venous return Reverse process of cardiac failure (mediastinal compression)5. Neural tube defects Damage to exposed neural tube Prevention of exposure of the spinal cord to amniotic Chronic cerebrospinal ﬂuid leak, leading to ﬂuid; restoration of CSF pressure, correcting Arnold-Chiari malformation and Arnold-Chiari malformation hydrocephalus6. Cardiac malformations Critical lesions causing irreversible hypoplasia Reverse process by anatomic correction of restrictive or damage to developing heart pathologySurgery on the Placenta, Cord, or Membranes7. Chorioangioma High-output cardiac failure by arteriovenous Reverse process of cardiac failure and hydrops shunting fetoplacentalis by ablation or reduction of ﬂow Effects of polyhydramnios8. Amniotic bands Progressive constrictions causing irreversible Prevention of amniotic band syndrome leading to neurologic or vascular damage deformities and function loss9. Abnormal monochorionic Intertwin transfusion leads to Arrest intertwin transfusion, prevent or reverse twinning: twin-to-twin oligopolyhydramnios sequence, cardiac failure and/or neurologic damage, including transfusion hemodynamic changes; preterm labor and at the time of in utero death rupture of the membranes; in utero damage Prolongation of gestation to brain, heart, or other organs In utero fetal death may cause damage to co-twin Fetus acardiacus and discordant Cardiac failure of pump twin and Selective feticide: to arrest parasitic relationship, to anomalies consequences of polyhydramnios prevent consequences of in utero fetal death Serious anomaly raises question for To avoid termination of the entire pregnancy termination of pregnancy or selective feticideHistorically, in utero treatment of hydrocephalus was attempted but abandoned. In the late 1990s, indications 5 and 6 were added; 7 to 9 were typicalresults of the introduction of obstetric endoscopy in fetal surgery programs.TABLE 24-2 CRITERIA FOR FETAL SURGERY (Premium Poly CS 57, US Surgical, Norwalk, CT) to prevent intraopera- tive maternal hemorrhage. Location of the uterine incision largely1. Accurate diagnosis and staging is possible, with exclusion of depends on placental position, as determined by sterile ultrasound. associated anomalies The fetus is partially exposed, and sometimes exteriorized, and2. Natural history of the disease is documented, and prognosis monitored while the procedure is performed, using ultrasound, pulse- established. oximetry, or direct fetal electrocardiography.9 Additional analgesics,3. Currently no effective postnatal therapy.4. In utero surgery proven feasible in animal models, reversing atropine, and pancuronium or vecuronium are given to the fetus to deleterious effects of the condition. suppress the fetal stress response, bradycardia, as well as to immobilize5. Interventions performed in specialized multidisciplinary fetal it. The fetus is kept warm through the use of intrauterine infusion of treatment centers with strict protocols and approval of the Ringer’s lactate at body temperature, and intrauterine volume and local ethics committee, and with informed consent of the pressure are maintained as close as possible to physiologic levels. After mother or parents. completion of the fetal portion of the procedure, the uterus is closed in two layers with resorbable sutures, amniotic ﬂuid volume is restored,Adapted from Harrison MR, Adzick NS: The fetus as a patient: Surgicalconsiderations. Ann Surg 213:279-291; discussion, 277-278, 1991. and intra-amniotic antibiotics are administered. The hysterotomy is covered with an omental ﬂap. Postoperatively, the patient is managed
CHAPTER 24 Invasive Fetal Therapy 435in intensive care while receiving aggressive tocolysis with magnesium ﬁber-endoscopes rather than conventional rod lens scopes, and as thesulfate and, when required, additional agents. number of pixels has increased, image quality has improved markedly. Complications of open fetal surgery include preterm contractions, Working length must be sufﬁcient to reach all regions of the intrauter-maternal morbidity from tocolysis, rupture of membranes, and fetal ine space. Amniotic access is facilitated by thin-walled, semiﬂexibledistress. Postoperative uterine contractions are the Achilles’ heel of disposable, or larger-diameter reusable metal cannulas, so that instru-open fetal surgery, but new tocolytic regimes have improved tocolytic ment changes are possible. Once inside the amniotic cavity, the obtura-efﬁcacy while limiting maternal side effects. Amniotic ﬂuid leakage tor is replaced by the fetoscope. Technical handbooks provide detailsthrough the hysterotomy site (or, more commonly, vaginally because of use of these instruments and a discussion of distention media.3,4of membrane rupture) can occur. With signiﬁcant postoperative oligo- Instrument insertion is facilitated with local or locoregional anesthesia,hydramnios, delivery may be necessary because of fetal distress. In which is injected along the anticipated track of the cannula down torecent case series on myelomeningocele repair, patients left the hospital the myometrium.within a few days, a much shorter interval than previously.10,11 Delivery Despite the minimally invasive nature of fetoscopy, it continues toby cesarean section is mandatory to prevent uterine rupture. be associated with iatrogenic preterm premature rupture of mem- branes (pPROM) (Table 24-3). Initiatives that have been evaluated to treat or prevent this condition12 include attempts to repair defects withThe EXIT Procedure various tissue sealants applied either intracervically or intra-amnioti-The EXIT (ex utero intrapartum treatment) procedure is increasingly cally. These efforts have met with limited success, because fetal mem-used for selected fetal conditions and is an example of open fetal branes have limited ability to heal.13 The use of amniopatch as asurgery—in this instance, to establish functional and reliable fetal treatment modality for symptomatic iatrogenic pPROM was ﬁrstairway control while keeping the fetus attached to the uteroplacental described by Quintero in 1996,14 and since then a number of case seriescirculation by delivering only a portion of the fetus through a hyster- have been published15 for iatrogenic membrane rupture after amnio-otomy incision. EXIT is done under maximal uterine relaxation, so the centesis, fetoscopy, or chorionic villus sampling.16 Chorionic mem-maternal risks of this procedure are mainly hemorrhagic. Because of brane separation, without obvious amniorrhexis, is anotherthe complex interactions necessary between the anesthesia, obstetric, complication that may be treated with amniopatch.17and pediatric personnel, EXIT procedures require signiﬁcant advancepreparation and preassignment of the roles for the many physiciansand nurses involved. The indications for EXIT include congenital Fetal Pain Relief during Proceduresairway obstruction from laryngeal atresia, large head and neck tumors, Pain is a subjective experience occurring in response to impendingand malformations of the face and jaw. For more information on EXIT, or actual tissue damage. The subjective experience of pain requiresplease refer to the Online Edition of this chapter. nociception and an emotional reaction. Nociception requires an intact sensory system, whereas an emotional reaction requires some form of consciousness. It is difﬁcult to know the extent to which the fetus experiences pain. However, several indirect methods haveFetoscopy suggested that the fetus at least can feel pain. Anand and colleagues and Fisk and coworkers demonstrated that premature infants andInstrumentation fetuses display several stress responses during invasive procedures.18-20Fetoscopic procedures are minimally invasive interventions that can be These data indicate that the mid-gestational fetus responds to noxiousconsidered a cross between ultrasound-guided and formal surgical stimuli by mounting a distinct stress response as evidenced by an out-procedures. Fetoscopy must be organized so that the surgical team can pouring of catecholamines and other stress hormones as well as hemo-see simultaneously both the ultrasound and the fetoscopic image. dynamic changes. Consequently, management of fetal pain andSpeciﬁcally designed fetoscopes typically have deported eyepieces to associated stress response in utero during invasive fetal interventionsreduce weight and facilitate precise movements. Nearly all are ﬂexible is important.20 TABLE 24-3 RISK FOR PRETERM PREMATURE RUPTURE OF MEMBRANES (pPROM) AFTER FETOSCOPIC PROCEDURES Risk for pPROM Procedure (At Assessment) Diameter Instrument Reference Amniocentesis 1%-1.7% 22 ga (0.7 mm) Tabor et al, 1986; Eddleman et al, 2006 Amniodrainage 1% per tap 18 ga (1.2 mm) Umur et al, 2001 Mari et al, 2000 Cordocentesis 3.7% (<37 wks) 20 ga (0.9 mm) Tsongsong et al, 2001 Shunt 15% (thorax) 7 Fr (2.3 mm) Picone et al, 2004 (thorax) 32% (bladder) Freedman, 1996 (bladder) Fetoscopic laser 7% (<1 wk) 10 Fr (3.3 mm) Yamamoto and Ville, 2005 (twins: 6%) 45% (<37 wks) Lewi et al, 2006 Cord occlusion 10% (<4 wks) 10 Fr (3.3 mm) Robyr et al, 2005 FETO 20% (<32 wks) 10 Fr (3.3 mm) Jani et al, 2006 FETO, fetal endoscopic tracheal occlusion.
436 CHAPTER 24 Invasive Fetal Therapy oliguric oligohydramnios, with the deepest vertical pocket (DVP) being 2 cm, while the recipient twin presents with polyuric polyhy-Fetal Therapy for dramnios (DVP cutoff of 8 cm before 20 weeks’ gestation and 10 cm after 20 weeks).7 Although growth restriction is often present in theComplicated Monochorionic donor twin, it is not essential for the diagnosis of TTTS. In severe casesTwin Pregnancies of TTTS, ultrasound signs of congestive cardiac failure resulting from ﬂuid overload in the recipient include a negative or reverse a-wave inMonochorionic twins constitute about 30% of all twin pregnancies21 the ductus venosus, pulsatile ﬂow in the umbilical vein, tricuspidand by deﬁnition share a single placenta and nearly always have vas- regurgitation, and signs of hypovolemia or increased vascular resis-cular anastomoses interconnecting their circulations.22 Because of tance in the donor, with absent or reversed ﬂow in the umbilicaloften unequal placental circulatory districts as well as cross-connecting artery.placental vessels, monochorionic twins have substantially greater mor- The differential diagnosis includes monoamnionicity, discordantbidity and mortality than their dichorionic counterparts.23 These result growth, isolated polyhydramnios or oligohydramnios, and severefrom complications such as TTTS, twin reversed arterial perfusion intertwin hemoglobin differences at the time of birth. TTTS does occur(TRAP) sequence, and, in the event of single intrauterine fetal demise in monoamniotic pregnancies and is characterized by polyhydramnios(IUFD), acute exsanguination of the surviving twin into the vascular of the common amniotic cavity with discordant bladder sizes. However,space of the demised twin.24 monoamniotic twins can move freely and usually their umbilical cords are entangled, whereas in diamniotic twins with TTTS, the donor is usually stuck against the uterine wall. Severe discordant growth is alsoTwin-Twin Transfusion Syndrome often confused with TTTS, as the growth restricted twin may appearTTTS occurs in 8% to 9% of monochorionic twin pregnancies and stuck because of oligohydramnios, but the appropriately grown twinrepresents the most important cause of mortality. The complication invariably has normal amniotic ﬂuid or only a mild degree of polyhy-typically becomes clinically evident between 16 and 26 weeks of gesta- dramnios that does not fulﬁll the criteria of TTTS.tion.25,26 The pathology is usually explained by unbalanced circulatorysharing between the twins across placental vascular anastomoses. StagingThese anastomoses are denoted as arterioarterial (AA), venovenous TTTS is currently staged according to the Quintero staging system,30(VV) or arteriovenous (AV).27 AA and VV anastomoses are bidirec- which is based on relative amniotic ﬂuid volume, Doppler waveforms,tional anastomoses, whereas AV anastomoses are unidirectional and and the bladder status in the donor (Table 24-4). Although thehence may create imbalance in interfetal circulation, leading to TTTS. Quintero staging system predicts outcome, it better reﬂects manifesta-The artery and vein of an AV anastomosis can be visualized on the tions of disease rather than a time sequence, as it is clear that cases canplacental surface as an unpaired artery and vein that pierce the chori- progress directly from stage I to stage V, and TTTS can appear as stageonic plate at close proximity. Bidirectional AA anastomoses are believed III. Attempts are now being made to improve the current stagingto protect against the development of TTTS, as most non-TTTS mono- system by incorporating a cardiac function score with echocardio-chorionic placentas (84%) have AA anastomoses in contrast to TTTS graphic features.31,32 However, its additional value for predictingplacentas (24%). Although vascular anastomoses are an anatomic outcome or the choice of treatment remains to be demonstrated.33prerequisite for the development of TTTS, the pathogenesis of TTTSis probably more complex,28 involving vasoactive mediators produced Treatmentby both donor and recipient.29 The mortality of untreated mid-trimester TTTS is more than 80% because of extreme prematurity with labor or pPROM with polyhy-Diagnosis dramnios, or as a result of fetal demise of one or both twins fromThe diagnosis of TTTS is based on stringent sonographic criteria of cardiac failure in the recipient or poor perfusion in the donor. In viewamniotic ﬂuid and bladder ﬁlling discordance. In the donor, there is of the poor outcome, the general consensus is that treatment should TABLE 24-4 QUINTERO STAGING SYSTEM Time in Pregnancy DVP Recipient DVP of Donor <20 weeks* ≥8 cm <2 cm ≥20 weeks* ≥10 cm* <2 cm With Either Stage I Stage II Stage III Stage IV Stage V Bladder ﬁlling in donor Absent bladder ﬁlling Abnormal Doppler ﬁndings: Hydrops fetalis Intra-uterine in donor Absent/reversed EDF umbilical artery (donor); fetal death Reversed a-wave ductus venosus (recipient) DVP, deepest vertical amniotic ﬂuid pocket; EDF, end-diastolic ﬂow. Twin-twin transfusion syndrome cases should have a deepest pool of 8 cm on the recipient side and a deepest pool of less than 2 cm on the donor side. Classiﬁcation is further made by the ﬁlling status of the bladder in the donor (Stages I and II). Additional (Doppler) ultrasound features upgrade stage. *In European centers, most use a cutoff of 10 cm for gestation over 20 weeks. For earlier presentations than 18 weeks, cutoffs have not been agreed upon.
CHAPTER 24 Invasive Fetal Therapy 437be offered. Unfortunately, even with the latest treatment modalities, still be separated from the chorion, hampering amniotic access andthe risk for adverse outcome remains signiﬁcant and the pregnancy making the degree of oligohydramnios or polyhydramnios difﬁcult tomust be followed carefully regardless. The option of pregnancy termi- measure. After 26 weeks, fetoscopic laser coagulation remains a validnation should be part of patient counseling. treatment option and appears to be associated with less major neonatal morbidity than repeated amnioreduction.36 From 32 weeks onward, AMNIOREDUCTION elective preterm delivery should be considered if lung maturation can Serial amnioreduction was the ﬁrst procedure offered to reduce be documented.polyhydramnios and intrauterine pressure in the hopes of alleviating Preoperatively, a detailed ultrasound scan is performed for diseasecontractions and prolonging the pregnancy. Theoretically, amniore- staging and to exclude discordant anomalies. Prophylactic antibioticsduction might also improve fetal hemodynamics by reducing the and prophylactic tocolytics are used. Fetoscopy is performed percuta-amniotic ﬂuid pressure on the placental vessels. Amnioreduction is a neously through a 3- to 4-mm incision under local or regional anes-relatively simple procedure involving aspiration of amniotic ﬂuid via thesia. High-quality videoendoscopic hardware with an excellent lightan 18-gauge needle under local anesthesia until restoration of normal source, video camera, and monitor, and with speciﬁcally designed 20-amniotic ﬂuid volume can be measured sonographically. The main to 30-cm ﬁberoptic or rod lens fetoscopes with a diameter of 1 toshortcoming of amnioreduction is its failure to address the cause of 3 mm, is used. For laser coagulation, a neodymium-yttrium aluminumthe disease, because the vascular anastomoses remain patent. Further- garnet (Nd-YAG) laser (minimal power requirements, 60 to 100 W) ormore, even if amnioreduction can resolve or stabilize stage I or II a diode laser (30 to 60 W) with ﬁbers of 400 to 600 μm provide optimaldisease, it fails in one third of cases,33 and, after failed amnioreduction, efﬁcacy.subsequent laser coagulation may be hampered by intra-amniotic Next, the positions of the fetuses, umbilical cord insertions, andbleeding, membrane separation, or unintentional septostomy. placenta are mapped by ultrasound. Under ultrasound guidance, the cannula or fetoscopic sheath is inserted into the recipient’s sac. Pref- SEPTOSTOMY erentially, the site of the trocar insertion is remote from the donor’s Intentional puncturing of the intertwin septum (i.e., septostomy), sac to avoid the risk of unintentional septostomy, and the trocar iswith or without amnioreduction, has been suggested to have beneﬁcial aimed to achieve a 90-degree angle with the vascular equator, as thiseffects largely based on the rarity of TTTS in monoamniotic twins. It provides the best opportunity for optimal coagulation. The vascularis proposed that the donor may be able to restore blood volume and equator can usually not be visualized on ultrasound unless there is aimprove perfusion by swallowing amniotic ﬂuid. A randomized trial marked difference in echogenicity between the two placental districts.37comparing septostomy with amnioreduction found similar rates of The trocar is therefore optimally inserted halfway along the imaginarysurvival of at least one twin. However, patients undergoing septostomy line between the two cord insertions. Occasionally, vision is hamperedwere more likely to require only a single procedure.34 Nevertheless, by blood or debris, in which case amnioexchange with warmedseptostomy brings with it the potential risks of cord entanglement Hartmann’s solution (heated by a blood warmer or a special amnio-resulting from an iatrogenic monoamnionic state, and it makes laser irrigator) can improve visibility.coagulation of the vascular anastomoses for progressive disease techni- A systematic inspection of the entire vascular equator is performed.cally much more challenging. The placental insertion of the intertwin septum is easily identiﬁed as a thin white line on the chorionic surface, and anastomosing vessels SELECTIVE FETICIDE leaving the donor usually cross under the septum in the direction of This procedure is usually performed by umbilical cord coagulation the recipient. Anastomosing vessels can also be identiﬁed starting fromwith laser or bipolar energy and is associated with an overall survival the recipient’s or donor’s cord insertion. Arteries are distinguishablerate of the remaining twin of about 70% to 80%.8,9 The most important from veins, as they cross over the veins and have a darker color becausedrawback of this approach is its maximum survival rate of 50%. Also, of their lower oxygen saturation.38 Not uncommonly, it may be impos-it may be unacceptable for many parents to sacriﬁce one twin without sible, because of the position of the intertwin septum, placenta, fetus,obvious structural pathology, and it may not be easy to determine or other physical limitations, to determine whether vessels anasto-which twin has the highest risk for adverse outcome. Therefore, this mose. In these instances, these vessels are coagulated as well, as the aimtechnique should be reserved for cases with severe discordant anoma- is to separate the two fetal circulations completely.lies, with an inaccessible vascular equator, with pPROM of one sac, or Coagulation is performed at a distance of approximately 1 cm andwith imminent IUFD. ideally at a 90-degree angle, using a nontouch technique (Fig. 24-1), starting at one placental border and ﬁnishing at the other end. Recently FETOSCOPIC LASER COAGULATION the sequential approach selecting A-V anastomoses ﬁrst has been advo- Laser coagulation of the vascular anastomoses was ﬁrst reported in cated.39 Sections of 1 to 2 cm are coagulated with shots of about 3 to1990 by De Lia and colleagues,35 who described nonselective coagula- 4 seconds, according to the tissue response. The use of excessive lasertion of all vessels crossing the intertwin membrane, thus arresting the power levels should be avoided, as this may cause vessel perforationtransfusion of blood and vascular mediators from donor to recipient and fetal hemorrhage. Once all vessels are coagulated, the vascularand functionally making the placenta dichorionic. However, nonselec- equator is inspected once more to ascertain that all anastomoses havetive coagulation of all vessels along the intertwin membrane causes been fully coagulated and that ﬂow has not resumed. The procedure issigniﬁcant parenchymal placental damage and probably increases the completed by amnioreduction until normal amniotic ﬂuid volumeprocedure-related fetal loss. Most fetoscopic laser centers therefore (DVP, 5 to 6 μm) is measured by ultrasound.avoid coagulation of nonanastomosing vessels and instead perform With an anterior placenta, the recipient’s sac as well as the anasto-selective coagulation of all visible anastomosing vessels along the vas- mosing vessels may be much more difﬁcult to access. Instrumentscular equator. for anterior placentas have been developed, but it is still unclear whether Fetoscopic laser coagulation is usually performed between 16 and these improve performance. Nonﬂexible rod lens telescopes have been26 weeks of gestation. Before 16 weeks, the amniotic membrane may fabricated with angles of inclination up to 30 degrees, or with an associ-
438 CHAPTER 24 Invasive Fetal TherapyFIGURE 24-1 Schematic drawing of fetoscopic laser coagulation. When the placenta is posterior, the scope can be directly inserted throughthe sheath without using a cannula (left). When the placenta is anterior (right), a curved sheath and a ﬂexible cannula (inset) can be used. Thisallows a change of instruments. (Drawing by K. Dalkowski, and modiﬁed with permission of Endopress Karl Storz.)ated deﬂecting mechanism for the laser ﬁber.40 Thus far, similar out- AMNIOREDUCTION VERSUScomes have been reported for anterior and posterior placentas.7 LASER COAGULATION Postoperatively, the patient should remain in the hospital for 1 to The Eurofoetus trial,7 with 142 enrolled patients with stage I to IV2 days while daily ultrasound scans are performed to document fetal disease diagnosed between 15 and 26 weeks of gestation, demonstratedviability, amniotic ﬂuid volume, and changes in the phenotypic fea- that fetoscopic laser coagulation is currently the best available treat-tures of TTTS, particularly bladder ﬁlling and Doppler parameters. ment option for TTTS (Table 24-5). Compared with amnioreduction,Postoperative transient hydropic changes and absent or reversed a- use of the laser was associated with a signiﬁcantly higher likelihood ofwave in the ductus venosus may occur in the donor twin.41,42 For fetuses the survival of at least one twin to 28 days of life (76% versus 56%).with absent end-diastolic ﬂow preoperatively, reappearance of end- Also, the median gestational age at delivery was higher in the laser thandiastolic ﬂow is observed in 53%.43 Laser coagulation may equalize in the amnioreduction group (33.3 weeks versus 29.0 weeks’ gestation),previously discordant umbilical venous blood ﬂow between donor and with 42% and 69% women, respectively, delivering at less than 32recipient.44,45 Catch-up fetal growth has also been described. weeks. Also, infants in the laser group had a lower incidence of cystic Fetal complications after laser treatment include fetal demise, iso- periventricular leukomalacia (laser 6% versus amnioreduction 14%).lated severe intertwin hemoglobin discordance, and persistent TTTS. Importantly, the Eurofoetus trial demonstrated that more than half ofPostoperative single IUFD occurs in about 33% and double IUFD in severe cerebral lesions identiﬁed postnatally appear to have had an4% of pregnancies. Single IUFD seems to affect donor and recipient antenatal origin.24,49 A recent systematic review of the Eurofoetus ran-equally,46 and approximately 60% are diagnosed within 24 hours and domized controlled trial and two other observational studies conﬁrmed75% within 1 week. Persistent TTTS complicates up to 14% of pregnan- that laser coagulation appears to be more effective in the treatment ofcies with two surviving fetuses 1 week after laser treatment47 and appears TTTS, with less perinatal neurologic morbidity and mortality.50to be related to missed, large, unidirectional anastomoses.48 Possible Huber and coworkers51 similarly demonstrated, in a consecutivetreatment options include repeat laser with backup cord coagulation, series of 200 pregnancies, a signiﬁcant trend toward reduced survivalamnioreduction, and elective delivery. A number of other, rarer com- after fetoscopic laser treatment with increasing stage. Survival of bothplications have been described, most of them occurring after amniore- twins was 75.9% for stage I, 60.5% for stage II, 53.8% for stage III, andduction as well. Please also refer to the Online Edition of this chapter. 50% for stage IV. At least one twin survived in 93.1% of pairs at stage
CHAPTER 24 Invasive Fetal Therapy 439 TABLE 24-5 LASER COAGULATION VERSUS AMNIOREDUCTION Laser Amnioreduction (N = 72) (N = 70) P Value Gestational age at randomization (wks) 20.6 (2.4) 20.9 (2.5) ns Quintero stage at randomization Stage I 6 (8.3%) 5 (7.1%) ns Stage II 31 (43.1%) 31 (44.3%) ns Stage III 34 (47.2%) 33 (47.1%) ns Stage IV 1 (1.4%) 1 (1.4%) ns Number of procedures 1* 2.6 (1.9) — AFV drained per procedure (mL) or in total overall amniodrainages: median (range) 1725 (500-5500) 2000 (243-4000) ns 3800 (600-18,000) <.001 Pregnancy loss at or within 7 days of the initial procedure 8 (11.6%) 2 (2.9%) .10 Premature rupture of membranes at or within 7 days of the ﬁrst procedure 4 (5.8%) 1 (1.5%) .37 Premature rupture of membranes at or within 28 days of the ﬁrst procedure 6 (8.7%) 6 (8.8%) .98 Intrauterine death ≤7 days of the ﬁrst procedure† 16/138 (11.6%) 9/136 (6.6%) .23 At least one survivor at 6 months of life 55 (76.4%) 36 (51.4%) .002 No survivors 17 (23.6%) 34 (48.6%) One survivor 29 (40.3%) 18 (25.7%) Two survivors 26 (36.1%) 18 (25.7%) At least one survivor at 6 months stratiﬁed by stage Quintero stages I and II 32/37 (86.5%) 21/36 (58.3%) .007 Quintero stages III and IV 23/35 (65.7%) 15/34 (44.1%) .07 Gestational age at delivery—median (interquartile range) 33.3 (26.1-35.6) 29.0 (25.6-33.3) .004 Neonatal and infant death 12 (8.3%) 41 (29.3%) ≤24 hours after delivery 6 (4.2%) 26 (18.6%) 1 to 7 days after delivery 4 (2.8%) 6 (4.3%) 7 to 28 days after delivery 1 (0.7%) 5 (3.6%) 28 days or more after delivery 1 (0.7%) 4 (2.9%) Intraventricular hemorrhage (grades III and IV)‡ 2 (1.4%) 8 (5.7%) .10* Donor 2 (2.8%) 2 (2.9%) 1.0 Recipient 0 (0.0%) 6 (8.6%) .02 Cystic periventricular leukomalacia§ 8 (5.6%) 20 (14.3%) .02* Donor 2/72 (2.8%) 5/70 (7.1%) .27 Recipient 6/72 (8.3%) 15/70 (21.4%) .03 Baseline characteristics according to group. Results reported as number of pregnancies [n (%)]. AFV, amniotic ﬂuid valve. *Two patients had two laser procedures. † With number of fetuses as denominator (P value adjusted for clustering). ‡ Severe intraventricular hemorrhage was deﬁned as ventricular bleeding with dilation of the cerebral ventricles (grade III) or parenchymal hemorrhage (grade IV). § Cystic periventricular leukomalacia was deﬁned as periventricular densities evolving into extensive cystic lesions (grade III) or extending into the deep white matter evolving into cystic lesions (grade IV).I, 82.7% at stage II, 82.5% at stage III, and 70% at stage IV. The survival multiple fetuses share a single placenta with multiple vascular anasto-rate of donors (70.5%) was similar to that of recipients (72.5%). moses, selective feticide by intravascular injection of potassium chlo-Hecher and associates demonstrated the importance of the learning ride may embolize the healthy fetus. Fetoscopic umbilical cord ligationcurve in laser photocoagulation, with increased experience leading to has been largely abandoned because it is a cumbersome and lengthyimproved survival,52 later gestational age at delivery, and a decrease in procedure, although it achieved an immediate and complete cordneurodevelopmental impairment.53 occlusion.56 With regard to long-term follow-up, a study of surviving infants(aged 14 to 44 months) from a laser photocoagulation series laserdemonstrated neurologic problems in 22% of survivors,54 of which Selective Feticide for11% were mild and 11% were severe. In a later series by the same Other Complicationsgroup, 7% of infants showed minor and 6% showed major neurologic For fetuses with TRAP sequence, needle-based coagulation techniquesabnormalities.53 The reduced neurologic impairment in the second using laser, monopolar, and radiofrequency energy57-59 have beenreport may be explained by increased operator experience. These adopted, involving the insertion of a 14- to 17-gauge needle into theresults are signiﬁcantly better than the 16% minor and 26% major acardiac twin’s abdomen under ultrasound guidance, and aiming forabnormalities in a cohort treated with amnioreduction.55 the intra-abdominal rather than umbilical vessels. The largest series, Selective feticide may be indicated for cases with a severe discordant of 29 monochorionic multiple pregnancies treated with radiofrequencystructural or chromosomal anomaly, with severe discordant growth energy between 18 and 24 weeks, reported a survival rate of 86%.59and a high risk for IUFD, or with TRAP sequence, and it may be indi- Median gestational age at delivery was 38 weeks (range, 24 to 40 weeks).cated for selected cases with TTTS. However, because monochorionic In another recent smaller series60 of 13 cases treated with radiofre-
440 CHAPTER 24 Invasive Fetal Therapyquency energy between 17 and 24 weeks, the survival rate was 94%. was reported by Hecher and coworkers (N = 60) with an 80% survivalAll patients delivered after 32 weeks, except for one patient compli- rate.76cated by pPROM at 26 weeks. More discussion of the techniques of management of TTTS, together with ﬁgures, tables, and references, can be found in the OnlineUmbilical Cord Occlusion Edition of this chapter.At present, laser or bipolar coagulation of the umbilical cord isour preferred approach and can be used for all indications from 16weeks onward.48,61 Early on, a double-needle loaded with a 1-mm feto-scope and a 400-μm laser ﬁber is used.62 Lasering of the cord allows Isolated Congenitaloptimal visual control but may fail beyond 20 weeks because of theincreasing size of the umbilical cord vessels.63,64 Ultrasound-guided Diaphragmatic Herniabipolar cord coagulation was therefore introduced for later gestational Congenital diaphragmatic hernia (CDH) occurs sporadically, with anages. incidence of 1 in 2500 to 1 in 5000 newborns. The term congenital Bipolar coagulation is performed with a 2.4- or 3-mm reusable diaphragmatic hernia designates a range of lesions, and outcomes areor disposable forceps. Under ultrasound guidance, a portion of the accordingly diverse.77 Ultimately, all the phenotypes of CDH arise fromumbilical cord is grasped at a convenient location and coagulation genetic mutations in one or several developmental pathways commoncurrent is applied in progressive increments until the appearance of for tissues of the fetal diaphragm and adjacent organs.turbulence and steam bubbles indicates tissue coagulation. Conﬁrma- Eighty-four percent of lesions are left-sided, 13% are right-sided,tion of arrest of ﬂow distal to the occlusion is performed by color and 2% are bilateral. Complete diaphragmatic agenesis, with hernia-Doppler. Even if there is no longer any visible ﬂow, two additional tion of the central tendineus portion and eventration, are other rarecord segments (preferably at a site more proximal to the target fetus) manifestations. Associated anomalies are present in 40% of cases,are coagulated. After completion of the coagulation procedure, which confers an increased risk for neonatal death, and less than 15%amnioreduction of excessive ﬂuid is carried out before removal of the of infants in this group survive.78,79 In the majority, however, CDH iscannula. The survival rates of umbilical cord coagulation in monocho- an isolated defect. Although CDH is in essence a defect in the dia-rionic twins is approximately 80%. In dichorionic and monochorionic phragm, the abnormal lung development that ensues confers its clini-triplets, the technique resulted in similar survival rates of 79%. About cal impact. CDH lungs have a reduced number of alveoli, thickenedhalf of the losses are attributable to intrauterine demise of the healthy alveolar walls, increased interstitial tissue, and markedly diminishedco-twin and about half to postnatal losses related to the very preterm alveolar air space and gas-exchange surface area. The conductingbirth, mostly related to pPROM.47,48 airways and associated blood vessels are diminished as well. Both lungs are typically affected, the ipsilateral more than the contralateral. ThereTwin Reversed Arterial Perfusion Sequence may be other anatomic aberrations present in the diaphragm and inAn extreme manifestation of TTTS is the TRAP sequence, which com- the upper gastrointestinal tract, such as the position of the liver, lowerplicates about 1% of monochorionic twin pregnancies. In the TRAP esophagus, and stomach.sequence, blood ﬂows from an umbilical artery of the pump twin in areversed direction into the umbilical artery of the perfused twin, viaan AA anastomosis. The perfused twin’s blood supply is by deﬁnition Prenatal Diagnosis of Congenitaldeoxygenated and results in variable degrees of deﬁcient development Diaphragmatic Herniaof the head, heart, and upper limb structures. Two criteria are neces- The diagnosis of CDH is usually made in the prenatal period whensary for the development of a TRAP sequence: an AA anastomosis and cystic masses are visualized in the chest or when cardiac deviation isa discordant development65 or intrauterine death of one twin,66 allow- noted in the axial view of the thorax (Fig. 24-2). Left-sided CDH typi-ing reversal of blood ﬂow. cally appears with rightward shift of the heart and the echolucent The increased burden to perfuse the parasitic twin puts the pump stomach and intestines in the left chest. Right-sided CDH is more dif-twin at risk for congestive heart failure and hydrops.67 Because of the ﬁcult to diagnose because the echogenicity of liver is similar to that ofrarity of the disorder, the natural history of antenatally diagnosed cases the mid-trimester fetal lung. The diagnosis is suggested when the fetalis still poorly documented, with reported survival rates for the pump heart is shifted farther into the left chest, or when color Doppler inter-twin varying between 14%68 and 90%.69 Data on long-term outcome rogation of the umbilical vein and hepatic vessels are shown to crossare not available, although the risk for cardiac and neurodevelopmen- the diaphragmatic boundary. The differential diagnosis includes cystictal sequelae may be high as a result of vascular imbalances in utero.70,71 or mixed masses (cystic adenomatoid malformation, bronchogenic,Several factors have been suggested to indicate a poor prognosis, such enteric and neuroenteric cysts, mediastinal teratoma, and thymicas a high ratio of the weight of the acardiac twin to that of the pump cysts), bronchopulmonary sequestration, and bronchial atresia. Intwin,72 a rapid increase in the acardiac mass,73 and small differences in these conditions, abdominal organs are not displaced into the chest.the umbilical artery Doppler values.74,75 These parameters were, When CDH is suspected, the patient should be examined forhowever, mostly studied in the late second and third trimesters and do associated cardiac, renal, central nervous system, and gastrointestinalnot necessary apply in the early second trimester, where spontaneous anomalies.80 Chromosomal anomalies are increased in CDH, so amnio-resolution as well as sudden death of the pump twin remain unpredict- centesis and genetic consultation are mandatory. For management,able. Early intervention is an option, as the diagnosis is now usually patients should be referred to tertiary centers accustomed to managingmade at an early stage in pregnancy, although the pump twin may complex congenital anomalies, in prenatal and in postnatal periods, tosurvive without any intervention in at least half of cases. For later optimize the necessary comprehensive diagnostic and prognosticprocedures, umbilical cord coagulation as well as needle-based intrafe- assessments, provide counseling on which parents can base furthertal coagulation techniques are both suitable treatment options. The decisions, and eventually offer timed delivery followed by optimallargest experience with fetoscopic laser coagulation for this indication postnatal care.
CHAPTER 24 Invasive Fetal Therapy 441 A B C D FIGURE 24-2 Ultrasound of a fetus with congenital diaphragmatic hernia. A, Measurement of the lung-to-head ratio (LHR) in a section through the four-chamber view, with the longest-axis method and the tracing method. B, Measurement of LHR, 1 day after balloon insertion. Echogenicity has changed. C, Herniation of the liver. D, Visualization of the major vessels helps identify the liver position. of outcome.86 Because the echogenicity of liver is comparable to thatPrognostic Indicators for Congenital of the lung, Doppler interrogation of the umbilical vein and hepaticDiaphragmatic Hernia vessels helps in determining its position above or below theMultiple prognostic measures have been proposed, including propor- diaphragm.tions of the cardiac ventricles, amniotic ﬂuid volume, degree of medi- From 12 to 32 weeks, lung area in the fetus increases four timesastinal shift, and position of the stomach. However, most have not been faster than head circumference, so the LHR must be referenced toshown to reliably provide clinically useful correlations. Because the gestational age.35 The gestational effect on LHR can be accommodatedmost critical problems of the neonate with CDH are lung hypoplasia by expressing the observed LHR as a ratio of the expected mean forand pulmonary hypertension, quantitation of relative pulmonary mass gestation. A study from the CDH antenatal registry (354 fetuses) withusing a variety of imaging techniques has been most successful.81 unilateral isolated CDH evaluated these measures between 18 and 38 The best validated measurement is the ratio of lung area (assessed weeks.85 Observed-to-expected (O/E) LHR predicted outcome well,by two-dimensional ultrasound through the contralateral lung) to regardless of the gestational age at measurement, and also correlateshead circumference (LHR).82 Although different methods for measur- with short-term morbidity indicators (Fig. 24-3).87ing the LHR have been proposed, the most reproducible and accurateinvolves tracing the lung contours (see Fig. 24-2).83 The predictivevalue of LHR was validated in 184 consecutive cases of isolated left- Prediction of Pulmonarysided CDH; the fetuses were examined at 22 to 28 weeks of gestation Arterial Hypertensionand were born live beyond 30 weeks at 10 centers (Table 24-6).83-85 In utero assessment of lung vasculature can be obtained by measuringLiver herniation has also been shown to be an independent predictor the number of branches and vessel diameters, and by performing mea-
442 CHAPTER 24 Invasive Fetal Therapy TABLE 24-6 NEONATAL OUTCOME AS A FUNCTION OF LHR IN FETUSES WITH LEFT-SIDED ISOLATED CDH AND LIVER HERNIATION, EXPECTANTLY MANAGED VERSUS AFTER FETO Degree of Pulmonary Hypoplasia LHR N Expectant Management* LHR N FETO† Extreme 0.4-0.5 2 0 (0%) 0.4-0.5 6 1 (16.7%) Severe 0.6-0.7 6 0 (0%) 0.6-0.7 13 8 (61.5%) 0.8-0.9 19 3 (15.8%) 0.8-0.9 9 7 (77.8%) LHR < 1.0 27 3 (11.1%) LHR < 1.0 28 16 (57.1%) Moderate 1.0-1.1 23 14 (60.9%) 1.0-1.1 na 1.2-1.3 19 13 (68.4%) 1.2-1.3 na Mild 1.4-1.5 11 8 (72.7%) 1.4-1.5 na ≥1.6 6 5 (83.3%) ≥1.6 na total 86 43 (50%) *Data from Kinsella J, Parker T, Dunbar I, et al: Noninvasive delivery of inhaled nitric oxide therapy for late pulmonary hypertension in newborn infants with congenital diaphragmatic hernia. J Pediatr 142:397-401, 2003. † Data from Jani J, Nicolaides KH, Gratacos E, et al, and the FETO task group: Fetal lung-to-head ratio in the prediction of survival in severe left-sided diaphragmatic hernia treated by fetal endoscopic tracheal occlusion (FETO). Am J Obstet Gynecol 195:1646-1650, 2006. CDH, congenital diaphragmatic hernia; FETO, fetal endoscopic tracheal occlusion; LHR, fetal lung-to-head circumference ratio; na, not applicable because these fetuses were not eligible for FETO.surements of ﬂow and resistance with two- or three-dimensional tech- moderate cases, the predicted survival rate is more than 60%, soniques. The Toronto group demonstrated that the diameter of the planned delivery at a referral center is appropriate. In more severeipsilateral branch of the main pulmonary artery is related to the sever- cases, especially with associated anomalies, other options should beity of hypoplasia in the prenatal as well as in the postnatal period, discussed, including termination of pregnancy.where it is also a signiﬁcant predictor of morbidity.88-90 Recently, Ruano The patients with a poorer prognosis are also candidates for anand colleagues established nomograms for main pulmonary artery antenatal intervention that could reduce the likelihood of lethal pul-branch diameters, which will allow proper validation of this concept.91 monary hypoplasia. Previously, antenatal therapy consisting of in uteroThey also proposed three-dimensional power Doppler to assess the anatomic repair using hysterotomy and direct fetal surgery was advo-entire lung vasculature, and to predict neonatal survival and the occur- cated, but this has been abandoned because of poor results. Subse-rence of pulmonary hypoplasia.92 quently, Di Fiore and colleagues revived the concept of triggering lung growth by tracheal occlusion,98 based on the observation that fetuses with congenital high airway obstruction syndrome (CHAOS) displayPostnatal Management of Congenital impressive lung growth, probably because of the lung-distending pres-Diaphragmatic Hernia sure of the trapped pulmonary secretions. However, lung growth aloneThe ill development of the fetal lung leads to variable degrees of respi- is not adequate for postnatal pulmonary function, as lung epithelialratory insufﬁciency and pulmonary hypertension after birth. The aber- maturation is also necessary. Under normal circumstances, fetal breath-rant vasculature is also more sensitive to hypoxic vasoconstriction, ing movements promote ﬂuid ﬂow out of the airways, creating cyclesleading to pulmonary hypertension and further increasing right-to-left of tissue stretch and relaxation. It is now clear that breathing move-shunt. This creates a vicious cycle that prevents gas exchange of the ments are important for an appropriate balance between growth andshunted blood and increases acidosis and hypoxia. Pulmonary hyper- differentiation.99tension is increasingly being treated by inhaled nitric oxide.93 Admin- Temporary tracheal occlusion takes advantage of this principle, itsistration of prostaglandin E1 has been advocated to keep the ductus beneﬁcial effects being a function of the timing and duration of thearteriosus patent in cases with severe secondary left ventricular cardiac occlusion. When sustained until birth, lung growth is vigorous butdysfunction.94,95 Survival statistics for postnatally managed CDH are airway epithelial maturation is compromised. Experimental models,optimized in specialized centers: in France, referral centers have higher now conﬁrmed with clinical experience, demonstrate that in uterosurvival rates (41% to 66%; P = .03)96; in Canada, high-volume centers placement, then reversal of occlusion (plug-unplug sequence) achieves(>12 CDH admissions over the 22-month period) with centralized improved lung volume and maturation.100 Antenatal tracheal occlusionmanagement report a 13% higher survival rate than low-volume is clinically applied at 26 to 28 weeks under locoregional anesthesiacenters.97 using fetoscopy to insert a balloon into the fetal trachea. The occlusion is reversed by removing the balloon fetoscopically at 34 weeks to allow fetal breathing and epithelial maturation.101,102 If preterm labor pre-Antenatal Management for Congenital cludes fetoscopic removal, emergency peripartum removal by laryngo-Diaphragmatic Hernia tracheoscopy or an EXIT procedure may be required.103Meticulous preparation for a possible antenatal surgical procedure for A recent report of a randomized controlled trial by Harrison andCDH is critical. Informed consent requires counseling that includes coworkers showed that prenatal tracheal occlusion did not increasedescribing the typical postnatal course of a newborn with CDH, survival.104 In that study, however, fetuses with O/E LHR up to 36%together with the range of expected morbidities that might be encoun- (LHR = 1.4) were offered fetal therapy. It is therefore not surprisingtered. Counseling should also include individualized information that postnatal management yielded results equivalent to those of ante-derived from the imaging assessments (O/E LHR < 25%). In mild or natal surgery. There were not enough cases with extreme or severe
CHAPTER 24 Invasive Fetal Therapy 443 6 160 Lung area to head circumference ratio (LHR) 140 5 Observed/expected LHR (%) 120 4 100 3 80 60 2 40 1 20 0 0 12 16 20 24 28 32 12 16 20 24 28 32 A Gestation (wks) B Gestation (wks) 90 90 80 80 70 Observed/expected LHR (%) 70 60 60 Survival (%) 50 50 40 40 30 30 20 20 10 10 0 0 18 22 26 30 34 38 25 26-35 36-45 46-55 55 C Gestation (wks) D Observed/expected LHR (%) FIGURE 24-3 Ratio of lung area to head circumference (LHR). A, Measurements of right LHR in normal fetuses throughout pregnancy. The curve shows LHR as a function of gestational age. B, Plot of the observed- to-expected (O/E) LHR with the mean and the 95th and 5th percentiles in normal fetuses. C, Right O/E LHR in fetuses with isolated left-sided diaphragmatic hernia. The shaded area depicts the lower part of the normal range. Closed circles (observations), solid line (mean), nonsurvivors; open circles (observations), dotted line (mean), survivors. D, Survival rates according to the O/E LHR in fetuses with isolated left-sided diaphragmatic hernia and liver herniation (N = 161). (A and B from Peralta CF, Cavoretto P, Csapo B, et al: Assessment of lung area in normal fetuses at 12-32 weeks. Ultrasound Obstet Gynecol 26:718-724, 2005; C and D modiﬁed from Jani J, Nicolaides KH, Keller RL, et al: Observed to expected lung area to head circumference ratio in the prediction of survival in fetuses with isolated diaphragmatic hernia. Ultrasound Obstet Gynecol 30:67-71, 2007.)hypoplasia. These are the subjects who are offered fetal endoscopic cases, patients are managed expectantly in the hospital so that the tra-tracheal occlusion (FETO) in the current European programs, where cheal balloon can be promptly removed once labor or chorioamnion-the presence of liver herniation and O/E LHR < 27% to 28% (corre- itis occurs. Thus far, more than 75% of patients have delivered beyondsponding to LHR < 1.0 in the early third trimester) are strict criteria 34 weeks (mean gestational age at birth, 36 weeks), signiﬁcantly latertoday. than the 31 weeks observed by Harrison and colleagues.104 Neonatal The European FETO task force has reported a survival rate of 50% survival rate is higher with prenatal than with perinatal balloonto 57% in this group of fetuses with otherwise poor predicted sur- retrieval (83.3% versus 33.3%; P = .013), a trend persisting until dis-vival.105 Iatrogenic pPROM remains a major complication. In such charge (67% versus 33%; NS). Major predictors of survival are gesta-
444 CHAPTER 24 Invasive Fetal Therapytional age at delivery and lung size prior to FETO.74 Similarly, the with pulmonary lymphangiectasia, and the prognosis remains poorincrease in lung area or volume after FETO is also an independent despite fetal treatment.predictor of survival.106,107 Trials are currently under design in The principal indication for fetal intervention for pleural effusionsEurope. is fetal hydrops, because the survival rate is 30%, compared with 80% when there are effusions without fetal hydrops.113 Thoracoamniotic shunting is performed using a double-pigtail catheter or via serial thoracentesis.116 There is no proof that serial puncture is better thanCongenital Cystic shunting, and the complication rate of this procedure is about 15%.Adenomatoid Malformation Iatrogenic pPROM117 and procedure-related fetal loss occurs in 5% to 10%. Shunt dislodgment has been described, but posterior insertionCongenital cystic adenomatoid malformation (CCAM) is a common may prevent the fetus from pulling the shunt out. Preterm birth ispulmonary malformation estimated to occur in approximately 1 in common, with a mean gestational age of 34 to 35 weeks.3000 to 5000 pregnancies. CCAM is a dysplastic or hamartomatoustumor with overgrowth of terminal bronchioles and reduction innumber of alveoli. Patients usually present with a thoracic massinvolving only one pulmonary lobe. In approximately 40%, the CCAM Lower Urinary Tractmay have a systemic vascular supply, similar to bronchopulmonarysequestration (BPS), and these forms are deﬁned as hybrid CCAM- ObstructionBPS. The classic pathologic classiﬁcation as proposed by Stockerestablished three types.108 Chapter 18 gives details of diagnosis and Pathophysiologyprognosis. Congenital anomalies of the genitourinary tract are the most com- When the CCAM is larger, there is an increased risk for pulmonary monly identiﬁed during prenatal ultrasound screening, with an inci-compression and hydrops. Use of the ratio of the mass area to head dence of up to 0.4%. Lower urinary tract obstruction (LUTO) is acircumference (CCAM volume ratio [CVR]) has been proposed as a descriptive term for a heterogeneous group of relatively common con-gestational-age-independent prognostic measure.109 CCAM volume ditions (1 in 5000 to 8000 male newborns). Posterior urethral valves is(in milliliters) is measured using the formula for an ellipse (length × by far the most common cause (at least one third in autopsy series),height × width × 0.52). When CVR is higher than 1.6, an 80% risk for but other conditions such as stenosis of the urethral meatus, anteriorfetal hydrops is predicted. When the CVR is high but fetal age is greater urethral valves, urethral atresia, ectopic insertion of a ureter, andthan 34 weeks, respiratory distress should be anticipated, and the EXIT (peri)vesical tumors are included.procedure with lobectomy in a tertiary care center should be consid- LUTO leads to bladder distention with compensatory hypertrophyered.110 Hydrops after viability should prompt delivery. In the previable of the smooth muscle of the bladder wall. Over time, bladder compli-period, fetal intervention can be lifesaving, but no large series is avail- ance and elasticity decrease and may contribute to poor postnatalable to judge among options. Percutaneous puncture and thoracoam- bladder function. Elevated bladder pressure inhibits urinary inﬂowniotic shunting of macrocystic masses has been reported and has the from above, resulting in reﬂux hydronephrosis. Progressive pyelectasisadvantage of minimal invasiveness.111 Wilson recently reviewed experi- and calyectasis compress the delicate renal parenchyma, leading toence with 23 shunted cases at a mean gestational age of 21 to 22 functional abnormalities in the medullary and eventually the corticalweeks.112 The mean CVR in this group was 2.4, which fell to 0.7 after regions, leading eventually to renal insufﬁciency. Concurrently, amni-shunting. The mean interval from shunt to delivery was 11.8 weeks otic ﬂuid volume falls, and pulmonary hypoplasia evolves. This condi-(36.3 weeks at birth). The overall survival was 74%, with one fetal and tion is reproducible in animal models, and more importantly, reversalﬁve neonatal deaths, correlating with a shorter shunt-to-delivery inter- of the obstruction both experimentally and clinically leads to reaccu-val. A recent systematic review concluded that CCAM cyst drainage mulation of amniotic ﬂuid.118improves perinatal survival among hydropic fetuses.113 For solid lesions, lobectomy via open fetal surgery can be consid-ered. In a series of 22 cases receiving surgery between 21 and 31 weeks, Case Selectionthere were 11 long-term survivors who were developmentally normal Because LUTO can result in neonatal lethal pulmonary hypoplasia and(up to 12 years of age).114 Hydrops resolved in 1 to 2 weeks, followed renal failure, careful selection of candidates for antenatal interventionby normalization of the mediastinum, and the remaining lung under- is paramount, ensuring that procedures to relieve LUTO are offeredwent impressive catch-up growth. Causes of fetal death despite fetal only to fetuses with sufﬁcient renal function.119 Because fetal renalsurgery were termination of pregnancy for Ballentyne syndrome (one function cannot be determined on a single urine sample, the best pre-patient), preterm labor and/or chorioamnionitis (two patients), and diction is obtained by two or more sequential vesicocenteses severalfetal hemodynamic compromise leading to intraoperative death in six days apart.120 The commonly recognized prognostic thresholds arefetuses and postoperative death in another two. shown in Table 24-7. The prenatal evaluation of fetuses with the sonographic ﬁndings of LUTO must be comprehensive, and coexisting structural and chromo-Other Thoracic Lesions somal anomalies must be excluded before intervention can be consid-Pleural effusion or fetal hydrothorax is a typical symptom of other ered.121 Female fetuses very often have more complex syndromes ofpathology. Pleural effusions have the potential to cause mediastinal cloacal malformations that may not beneﬁt from in utero shuntshift, abnormal venous return and secondary lung compression, therapy. Because of the presence of oligohydramnios or anhydramnios,hydrops, and intrauterine demise. Most primary effusions are lym- it may be necessary to obtain karyotype by transabdominal chorionicphatic in origin, but they may be associated with other anomalies villus or fetal blood sampling and thorough structural assessment of(25%), including aneuploidy (7%).115 Bilateral effusions are associated the fetus after amnioinfusion.
CHAPTER 24 Invasive Fetal Therapy 445 TABLE 24-7 FETAL URINE ANALYSIS of patients for performing such interventions remains unclear. Fur- PROGNOSTIC THRESHOLDS thermore, in utero fetal cardiac intervention is still frequently bound by technical limitations, the often late timing of diagnosis, and our Electrolytes Good Prognosis Poor Prognosis current insufﬁcient understanding of these diseases. Sodium <90 mmol/L >100 mmol/L Chloride <90 mmol/L >100 mmol/L Antenatal Intervention for Congenital Osmolality <180 mOsm/L >200 mOsm/L Total protein <20 mg/dL >40 mg/dL Heart Defects β2-microglobulin <6 mg/L >10 mg/L Fetal valvuloplasty is the most frequent intrauterine intervention, with indications including management of critical aortic and pulmonary stenosis or atresia, atrial septostomy for highly restrictive foramen ovale with aortic stenosis, and hypoplastic left heart syndrome (HLHS). Table 24-8 summarizes experiences with conditions that are currentlyOutcomes for Antenatal Therapy of claimed to be candidates for prenatal intervention, as recently sum-Lower Urinary Tract Obstruction marized by Matsui and Gardiner.128Vesicoamniotic shunts bypass the urethral obstruction, diverting the In utero valvuloplasty is typically done by a team of fetal medicineurine into the amniotic space and allowing drainage of the upper specialists familiar with intrauterine invasive procedures, and interven-urinary tract and prevention of pulmonary hypoplasia by restoration tional pediatric cardiologists. Levine and Tworetzky recently reviewedof amniotic ﬂuid volume. Initial experience showed good outcomes.122 their experience with this procedure.129 External version is performedIn the largest and longest-term documented experience from Biard and as required, and the fetus is positioned with the left chest facing ante-coworkers,123 postnatal diagnosis of the type of urinary obstruction riorly and the insertion track free of limbs or cord. Achieving thesewas highly predictive of long-term renal outcome. Fetuses with poste- optimal but essential conditions sometimes requires a laparotomy.rior urethral valves had better outcomes than those with urethral Fetal analgesia and immobilization are required; the mother can beatresia or the prune belly syndrome. Most children were developmen- sedated or more rarely is given general anesthesia. Bradycardia duringtally normal, but pulmonary problems may persist. Among 18 survi- the procedure occurs in 9% and hemopericardium in less than 1%,vors, six had acceptable renal function, four had mild insufﬁciency, and and the fetal loss rate has been reported as 3.8%.130 Under ultrasoundsix required dialysis and transplantation. guidance, the valve is dilated with a balloon up to 1.2 to 1.5 times the Other case series have had similar results, suggesting that even with size of the annulus.favorable pre-procedure urine proﬁles, up to half of survivors have HLHS is a particularly severe valvular abnormality amenable tochronic renal insufﬁciency in childhood.124,125 The systematic review by prenatal intervention. The range of HLHS structural malformationsClark and coworkers124 indicated that there was a lack of high-quality includes critical aortic stenosis, unbalanced atrioventricular septalevidence to reliably support the clinical practice of vesicoamniotic defects with hypoplasia of the left heart and aorta, severe aortal coarc-shunting despite an improvement in perinatal survival (odds ratio tation, and the association of atresia or hypoplasia of both the aortic[OR], 2.5; 95% conﬁdence interval [CI], 1.0 to 5.9; P < .03). Subgroup and mitral valve. Currently, the only available neonatal treatment isanalysis indicated that improved survival was most likely in fetuses staged palliative surgery (Norwood operation, followed by a Fontanwith a deﬁned poor prognosis (based on a combination of ultrasound procedure), with an overall 5-year survival of 70%. However, moreappearance and fetal urinary analytes) (OR, 8.0; CI, 1.2 to 52.9; than half will require further surgery before the age of 5 years.131P < .03). It has now been proposed to evaluated this therapy in a Balloon valvuloplasty for HLHS was ﬁrst technically successful inmulticenter randomized trial, referred to as a PLUTO trial.122 1991, and by 2000, 12 cases had been performed world-wide and com- piled. However, technical problems were frequent, with a 50% balloon rupture rate and a 75% incidence of bradycardia. Only two patientsCongenital Heart Defects survived, and one of them had a two-ventricle circulation after birth.132 However, the more extensive experience of the Boston group learnsThe ﬁrst fetal cardiac intervention was an attempt by Carpenter and what can be achieved after the learning curve. From the initial 20colleagues126 at fetal pacing for complete heart block because of circu- cases,133 there were 13 successful procedures between 21 and 29 weeks.lating anti-Ro/La antibodies. In 1991, Maxwell and coworkers127 per- In their most recent series, 28 of 38 treated fetuses survived. Neverthe-formed two percutaneous valvuloplasties, and since then, there has less, three died in utero after the procedure, one was born prematurely,been an increase in experience with minimally invasive treatment for and two patients opted for termination. Postprocedure assessments ofcongenital heart defects. Still, most centers only have sporadic experi- the left heart showed growth of the ventricle, of the mitral and aorticence, and only a few, such as Children’s Hospital in Boston, have valves, and of ascending aorta. Of the seven born live, three were surviv-achieved a wider experience.128 ing with biventricular circulation, and another had palliative surgery. Most congenital heart defects can be operated on postnatally with There are also other indications for in utero valvuloplasty and dis-low (<5%) mortality and good quality of life in survivors. However, ruption of the atrial septum. More information on these indicationswhen this is not the case, and when the abnormal cardiac anatomy and procedures for fetal cardiac diseases is available in the Onlinewould lead to progressive myocardial and pulmonary damage during Edition of this chapter.the pregnancy and would ultimately preclude effective postnatal treat-ment, timely fetal intervention can be lifesaving. Antenatal interven-tion theoretically reduces intraventricular pressure, improves coronaryperfusion, reduces ischemic damage, allows ventricular growth, and Myelomeningoceleavoids induction of myocardial ﬁbroelastosis, thus enabling improved Neural tube defects (NTD) are a major source of mortality and mor-functional postnatal repair. However, the exact timing and selection bidity. The pathophysiology and details of diagnosis are discussed in