Fetoscopic and Open Transumbilical Fetal Cardiac Catheterization in Sheep
Potential Approaches for Human Fetal Cardiac Intervention
Background Shortening the prenatal disease course of severe aortic and pulmonary stenoses by balloon valvuloplasty may diminish their postnatal expression. The purpose of this study in fetal sheep was to assess the feasibility of fetoscopic and open transumbilical fetal cardiac catheterization guided by fetal transesophageal echocardiography to provide alternative approaches for human fetal cardiac intervention.
Methods and Results We studied a total of nine fetal sheep (95 to 103 days of gestation; term=145 to 150 days) and performed transumbilical fetal cardiac catheterization by a minimally invasive fetoscopic (n=6) or an open fetal surgical approach (n=3). Monitored by fetal transesophageal echocardiography, with an 8F or 10F, 10-MHz intravascular ultrasound catheter we placed guidewires and interventional catheters via the umbilical arterial route into the fetal heart. In three of the fetuses, we created supravalvar pulmonary artery stenosis by open fetal cardiac surgery. After fetal and maternal recovery, we exteriorized these fetuses and performed open transumbilical fetal cardiac catheterization with successful pulmonary arterial angioplasty in two. Three fetuses survived fetoscopic transumbilical catheterization for 1 or 2 days and died most likely of blood loss after sheath dislodgment (n=1) or removal (n=2). By securing the sheath insertion site with a suture, we prevented sheath dislodgment and minimized bleeding during sheath removal in three fetuses. These fetuses then survived fetoscopic transumbilical fetal cardiac catheterization for 1 to 2 weeks before being killed.
Conclusions This study in fetal sheep demonstrates that fetoscopic and open transumbilical fetal cardiac catheterization are feasible and, guided by fetal transesophageal echocardiography, provide potential alternative approaches for human fetal cardiac intervention.
Echocardiographic observation of the prenatal course of congenital heart disease has prompted interest in performing fetal cardiac intervention in severe obstruction of the aortic or pulmonary valve.1 2 3 4 Both lesions can induce severe malformation and dysfunction in the obstructed ventricle and dependent cardiovascular structures, affecting postnatal treatment options and prognosis, and may require palliative rather than definitive surgical procedures.
Shortening the in utero disease course of fetal aortic and pulmonary valvar stenoses by balloon valvuloplasty or surgical valvotomy during fetal life is appealing because the severity of secondary cardiovascular injury may be reduced. Although open fetal cardiac surgery for valvotomy has not yet been achieved, fetal cardiac catheterization and balloon valvuloplasty have been performed in selected human fetuses.1 2 3 4 Cardiac access for this procedure has been obtained by ultrasound-guided direct puncture of the obstructed ventricle. For the successful performance of this technique, favorable fetal position, excellent imaging quality, and availability of sufficient acoustic windows are critical. Ultrasound-guided direct punctures of the fetal heart, however, have been associated with fetal hemopericardium, hemothorax, bradycardia, and death and may not be feasible in very young fetuses. In addition, during fetal balloon valvuloplasty, parts of the balloon catheters contacted the edge of the needle used for cardiac access and were torn or cut off of the main body of the catheter.1 2 4 Furthermore, repeated punctures of the heart were necessary to optimize positioning of interventional devices.2 These obstacles increase the risk for bleeding complications, embolic events, and injury to healthy myocardium as well as other cardiovascular and intrathoracic structures.
Alternatively, fetoscopic or open techniques allowing direct fetal access may offer a safer means for fetal cardiac catheterization than ultrasound-guided direct punctures of the heart.5 Fetal balloon valvuloplasty by itself would then resemble the postnatal approach. Fetoscopic and open procedures for fetal cardiac intervention, however, require imaging capabilities not attended by conventional fetal echocardiography that may interfere with the positioning of trocars and interventional devices or may not be feasible after gaseous distension of the uterine cavity.6 In this setting, fetal transesophageal echocardiography using an intravascular ultrasound catheter can be used for real-time imaging of fetal cardiac anatomy, function, and positioning of intravascular devices.6 7 The purpose of this study in fetal sheep was to assess the feasibility of fetoscopic and open transumbilical fetal cardiac catheterization guided by fetal transesophageal echocardiography to provide alternative approaches for fetal cardiac intervention.
We studied a total of nine Mixed Western breed fetal sheep (95 to 103 days of gestation; term=145 to 150 days). Each ewe was positioned supine, intubated, and ventilated with 100% oxygen and 0.5% to 2% halothane.
Fetoscopic Transumbilical Fetal Cardiac Catheterization
We placed three 5-mm trocars (Entec Corp; Origin Inc) into the amniotic cavity of each ewe either after maternal laparotomy (n=1) or by a percutaneous technique (n=5). Using fetoscopic instrumentation (Karl Storz), we inserted an 8F or 10F, 10-MHz intravascular ultrasound catheter (Boston Scientific Corp) into the fetal esophagus for transesophageal echocardiographic guidance.6 7 We then incised the umbilical cord sheath at its abdominal insertion site and dissected one of the two umbilical arteries. We punctured the dissected umbilical artery with an 18-gauge needle we had advanced percutaneously under fetoscopic visualization into the amniotic cavity. After successful umbilical arterial puncture, we inserted a 4F catheter sheath (Cook Inc) over a 0.038-in wire into the umbilical artery and advanced a 0.014-in straight, soft-tipped guidewire through the catheter sheath into the fetal abdominal aorta. Monitoring by fetal transesophageal echocardiography, we attempted placement of the guidewire into the great cardiac arteries and heart. We assessed acute complications related to catheter sheath placement as well as short-term survival (1 to 2 weeks). After the procedure, we performed daily maternal transabdominal ultrasound studies to observe fetal survival and assess umbilical blood flow. At autopsy, we examined the fetal heart and vessels as well as the sheath insertion site for perforations or damage to their endothelial linings from wire and catheter manipulations.
Open Transumbilical Fetal Cardiac Catheterization
We created a model of fetal supravalvar pulmonary artery stenosis by open fetal cardiac surgery in three fetal sheep between 75 and 93 days of gestation. We tied a band of polytetrafluoroethylene around the pulmonary artery and secured it with a single stitch of unresorbable suture material. After 1 to 3 weeks of fetal and maternal recovery, we exteriorized the fetuses by maternal laparotomy and hysterotomy and performed open transumbilical fetal cardiac catheterization between 96 and 103 days of gestation. We performed fetal transesophageal echocardiography and umbilical arterial dissection similar to the fetoscopic approach. We punctured the dissected umbilical artery with a 20-gauge needle and advanced a 0.014-in straight guidewire with a floppy tip through the needle into the fetal circulation. Guided by fetal transesophageal echocardiography, we placed the wire via the ductus arteriosus into the fetal right ventricle. We then removed the needle and inserted a 4-mm, 2-cm-long coronary artery balloon catheter (USCI Division) or a 4-mm, 0.9-cm-long coronary artery balloon catheter (Cordis Corp) over the wire into the umbilical artery. The balloon length was selected on the basis of the descending aorta–to–pulmonary valve distance. Using transesophageal echocardiographic guidance, we positioned the balloon catheter across the pulmonary artery band and hand-inflated the balloon three to five times to 4 to 5 atm pressure. After this procedure, we killed the fetuses with pentobarbital and assessed band integrity.
The study protocol was approved by the local committee on animal research and was performed according to institutional guidelines.
Fetoscopic Transumbilical Fetal Cardiac Catheterization
Using the fetoscopic approach, we successfully placed the catheter sheath into the umbilical artery in six fetal sheep (Fig 1⇓). During sheath insertion, we did not observe umbilical arterial constriction, and a mean time of 21 minutes (range, 15 to 35 minutes) was spent inside the fetal circulation before removal of all intravascular devices. Guided by fetal transesophageal echocardiography, we advanced the 0.014-in guidewire into the ascending aorta and across the ductus arteriosus into the pulmonary artery in all six fetal sheep. We placed the guidewire into the left ventricle and inflated a coronary artery balloon catheter above the aortic valve in three sheep (Fig 2⇓; the position of the wire is much better appreciated during real-time imaging).
All fetuses were alive at the end of the fetoscopic procedure as assessed by real-time fetal transesophageal echocardiography. Three fetuses survived fetoscopic catheterization of the umbilical artery and intravascular and/or intracardiac manipulation for 1 to 2 days, their early demise most likely being related to blood loss. In two of these fetuses, bleeding at the catheter sheath insertion site occurred after sheath removal. In the third fetus, bleeding occurred after the sheath was accidentally dislodged during intravascular balloon catheter manipulation. Bleeding stopped within 1 to 2 minutes by umbilical arterial constriction in these fetuses but was substantial because of the relatively large opening in the vessel wall.
By securing the sheath insertion site with a 5-0 polytetrafluoroethylene purse-string suture (Fig 1⇑), we prevented sheath dislodgment and minimized blood loss after sheath removal in three fetuses. These fetuses then survived fetoscopic transumbilical fetal cardiac catheterization for 1 to 2 weeks before being killed.
Open Transumbilical Fetal Cardiac Catheterization
Using the open approach, we inserted the guidewire and coronary artery balloon catheter into the umbilical artery in the three pulmonary artery–banded fetal sheep. We did not observe umbilical arterial constriction or any blood loss at the insertion site, and a mean time of 13 minutes (range, 10 to 15 minutes) was required for fetal pulmonary artery angioplasty. Fetal transesophageal echocardiography allowed accurate guidance of the balloon catheter across the pulmonary artery band in all three sheep. We successfully tore the band and dilated the banded lumen to 200% to 250% of the predilatation width in two of these fetuses (Fig 3⇓). In the third fetus, the suture material was too strong and could not be torn by hand inflation.
At autopsy, the fetal mean weight was 875 g (range, 550 to 1200 g). After fetoscopic or open transumbilical fetal cardiac catheterization, we did not observe vascular or ventricular perforation in any fetus. In one fetus that survived fetoscopic transumbilical cardiac catheterization for 2 weeks before being killed, the catheterized umbilical artery was completely obliterated by a 4-cm-long thrombus that originated from the repaired catheter sheath insertion site. In this fetus, the other umbilical artery was markedly dilated. In a second fetus that survived fetoscopic transumbilical cardiac catheterization for 2 weeks, both umbilical arteries were patent at autopsy. However, the catheterized right umbilical artery was significantly smaller than the noncatheterized left one. In this fetus, we observed increased pulsatility with low end-diastolic blood flow velocities in the catheterized artery by Doppler interrogation before termination. Conversely, pulsatility was decreased and end-diastolic flow velocities were increased in the noncatheterized artery.
Our study in fetal sheep demonstrates that fetoscopic and open transumbilical fetal cardiac catheterization are feasible and, guided by fetal transesophageal echocardiography, provide potential alternative approaches for human fetal cardiac intervention. Fetal transesophageal echocardiography permits accurate delivery of guidewires and interventional catheters via the umbilical arterial route into the fetal heart as well as dilatation of experimental lesions. The new approaches have the potential to facilitate balloon valvuloplasty in human fetuses and may be more reliable to obtain fetal cardiovascular access than currently performed ultrasound-guided direct punctures of the fetal heart.1 2 3 4 Those depend largely on excellent imaging quality, optimal alignment of interventional devices with the left or right ventricular outflow tract, and adequate fetal size and may not be feasible if acoustic windows are poor, the spatial orientation of the outflow tracts is unfavorable, or the fetuses are very young.
Fetoscopic Transumbilical Cardiac Catheterization
Fetoscopic transumbilical cardiac catheterization is more desirable than the open approach for human fetal cardiac interventions because it is less invasive for the mother and fetus, since laparotomy, hysterotomy, and fetal exteriorization are not required. Contrary to open transumbilical cardiac catheterization, however, the fetoscopic procedure is technically far more demanding and requires intensive training and collaboration of a skilled team because it encompasses a number of time-consuming steps that need to be performed under the restraints of laparoscopic procedures. These restraints are magnified eye-hand coordination, long instruments affecting surgical precision, and the two-dimensional imaging field lacking depth perception. Because of these restraints, the fetoscopic procedure is more time-consuming than the open approach. Yet the potential benefits of the fetoscopic approach, namely, reduced postoperative uterine irritation and the greatly decreased maternal operative trauma, make this approach worth exploring.
In our early experience, fetal bleeding at the time of sheath removal or by accidental sheath dislodgment was a serious complication of the fetoscopic approach. With increased familiarity with the procedure, however, we could minimize fetal blood loss by optimizing the fetoscopic setup and placing a purse-string suture into the arterial wall before sheath removal. These technical modifications minimized bleeding in three subsequent fetuses, which then survived fetoscopic transumbilical fetal cardiac catheterization for 1 to 2 weeks before being killed. Interestingly, umbilical arterial dissection as well as placement of the purse-string suture and sheath insertion did not result in umbilical arterial constriction in our study.
Complete thrombosis of the catheterized artery was seen in one of the fetuses that survived fetoscopic transumbilical cardiac catheterization. Although this complication was compatible with fetal survival in our short-term study, decreased fetal growth has resulted from long-term occlusion of one umbilical artery in sheep.8 Even though we performed systemic fetal heparinization, thrombosis of the distal umbilical artery might have been prevented by additional heparin injection in this area.
Open Transumbilical Fetal Cardiac Catheterization
Open transumbilical fetal cardiac catheterization after maternal laparotomy, hysterotomy, and fetal exteriorization can be performed more rapidly than fetoscopic transumbilical fetal cardiac catheterization because the time-consuming fetoscopic setup as well as sheath placement into the umbilical artery are avoided. With the open approach, both guidewire and interventional catheter can be inserted after simple needle puncture of the umbilical artery. Any bleeding at the insertion site can be easily controlled by arterial compression. We chose open transumbilical fetal cardiac catheterization for balloon dilatation in our pulmonary artery–banded fetuses because persistent oligohydramnios as well as decreased uterine compliance after the initial surgery for creation of the animal model did not allow us to create a favorable setup for fetoscopic fetal cardiac catheterization. These restraints to the fetoscopic approach should not be operative in human fetuses, which would not have been subjected to major surgery before fetal cardiac intervention.
Although open transumbilical fetal cardiac catheterization is rapid, permits comfortable control of the catheter insertion site, and can be performed when the fetoscopic approach is unmanageable, the effectiveness of open fetal cardiac procedures will be influenced by premature delivery resulting from uterine contractions after hysterotomy. The pharmacological suppression of these contractions remains the Achilles heel of current open surgical attempts in human fetuses.9 Conversely, the incidence of premature fetal delivery may be lower with fetoscopic techniques for fetal cardiac catheterization. In addition, fetoscopic techniques do not require cesarean section for fetal delivery, because the uterus is accessed through minimal openings. Cesarean section, however, is obligatory after open fetal surgery for the treated as well as for any future child because of the risk of uterine rupture during normal delivery.10
In conclusion, this study in fetal sheep demonstrates that fetoscopic and open transumbilical fetal cardiac catheterization are feasible and, guided by fetal transesophageal echocardiography, provide potential alternative approaches for human fetal cardiac intervention. Because it is less invasive and does not require maternal laparotomy, hysterotomy, and fetal exteriorization, the fetoscopic approach is more attractive. However, further studies are necessary to define and compare the physiological effects and long-term survival of the two approaches before they may be applied to the human fetus.
Dr Kohl is a research fellow supported by a research grant (Ko 1484/2-1) from the German Research Society (Deutsche Forschungsgemeinschaft), Bonn, Germany. We are grateful for the expert technical assistance of Vincente G. Lapiz. This study was made possible by grants from the Verein fu¨r das herzkranke Kind, Mu¨nster, Germany; Wilfried Vogeler, MD, Essen, Germany; and Microsurgery and the Operative Endoscopy Training (M.O.E.T.) Institute, San Francisco, Calif. We acknowledge the support of Boston Scientific Corp, Sunnyvale, Calif; Karl Storz America, Culver City, Calif; Cook Inc, Bloomington, Ind; USCI, Billerica, Mass; Cordis Corp, Miami, Fla; Entec Corp, Madison, Conn; and Origin Medsystems Inc, Menlo Park, Calif.
- Received June 5, 1996.
- Revision received September 12, 1996.
- Accepted September 30, 1996.
- Copyright © 1997 by American Heart Association
Maxwell D, Allan L, Tynan MJ. Balloon dilatation of the aortic valve in the fetus: a report of two cases. Br Heart J. 1991;65:256-258.
Chaoui R, Bollmann R, Goeldner B, Rogalsky V. Aortic balloon valvuloplasty in the human fetus under ultrasound guidance: a report of two cases. Ultrasound Obstet Gynecol. 1994;4:162A. Abstract.
Kohl T. Fetal cardiac intervention in congenital heart disease. In: Chervenak F, Kurjak A, eds. New Perspectives on the Fetus as a Patient. 2nd ed. New York, NY: The Parthenon Publishing Group; 1996:297-315.