- Email: [email protected]
Fetoscopic surgery for the treatment of congenital anomalies
Fetoscopic
Surgery for the Treatment
of Congenital
Anomalies
By James M. Estes, Thomas E. MacGillivray, Marc H. Hedrick, N. Scott Adzick, and Michael R. Harrison San Francisco, California l Fetoscopic techniques may broaden the indications for prenatal surgical intervention by obviating the risks of hysterotomy. For example, congenital obstructive uropathy has been treated by open vesicostomy and percutaneous catheter placement. The open approach is appropriate only for highly selected fetuses because of the inherent risks, whereas catheter drainage, though a safer procedure, is only useful for short-term therapy late in gestation due to frequent catheter obstruction and migration. The natural history of congenital obstructive uropathy mandates the need for improved therapy earlier in gestation, in order to salvage fetuses who would otherwise die of renal failure and pulmonary hypoplasia. We have developed a potential solution to this problem in which surgery is performed on the fetus without the risks of hysterotomy. Endoscopic fetal surgery uses a telescopic lens and operating instruments that are passed through small “ports“ in the uterus. A bubble of CO2 is used to displace amniotic fluid and provides excellent visualization in a magnified field. This approach is considerably less invasive than open fetal surgery and, therefore, is less likely to provoke preterm labor. In this study we corrected obstructive uropathy in midgestation fetal lambs using a new, expandable wire mesh stent that is placed endoscopically and should provide more reliable bladder drainage than existing catheters. The fetoscopic surgicaf approach can potentially expand the indications for in utero surgery by decreasing fetal risks, facilitating intervention earlier in gestation, and reducing preterm labor. As a consequence, the potential now exists to correct non-life-threatening malformations in utero. Copyright o 1992 by W. B. Saunders Company INDEX WORDS:
Fetoscopy;
obstructive
uropathy;
fetal sur-
gery.
I
NVASIVE techniques for prenatal fetal therapy are relatively limited. Open fetal surgery is an option for the highly selected fetus with a lifethreatening malformation, but carries significant morbidity predominantly from preterm labor.’ On the opposite end of the therapeutic spectrum are percutaneous techniques, which are used to drain fluid collections and place catheters. Failures due to tech-
From the Depanrnent of Surgery and the Fetal Treatment !+ograrn, University of California, San Francisco, San Francisco, CA. Third Place Winner in the Jens G. Rosencraru Resident Competition. Presented at the 43rd Annual Meeting of the Surgical Section of the Ametican Academy of Pediatrics, New Orleans, Louisiana, October 26-27, 1991. Address reprint requests to James M. Estes, MD, Fetal Treatment Program, Room HSE-585, University of California, 513 Pamassus St, San Francisco, CA 94143-0570. Copyright o 1992 by WE. Saunders Company 0022-3468/92/2708-0003$03.00/0
950
nical problems such as catheter obstruction and migration are common, limiting this procedure only to short-term therapy. The current management of congenital obstructive uropathy illustrates these problems well. It is a common malformation affecting approximately one of every one thousand fetuses, and causes death in the most severely affected.2 Fetal surgery has been successful with a few highly selected cases of this anomaly. The alternate treatments, which include percutaneous bladder stenting and repeated aspiration, offer only short-term palliation due to the high incidence of failure.3 Hydronephrosis is easily diagnosed in utero. Unilateral hydronephrosis is managed conservatively, while bilateral disease can present as a continuum of severity. Criteria have been developed to assess renal function in utero and these parameters include urine electrolytes, ultrasonographic appearance of kidneys (presence or absence of dysplastic cysts), urine output, and amniotic fluid volume.4 Several case reports in humans and experimental work in lambs demonstrate that in utero bladder drainage may restore amniotic fluid volume and prevent progressive renal deterioration and pulmonary hypoplasia,4J thus providing the scientific rationale for our study. Percutaneous transabdominal placement of a catheter is the most common procedure performed in utero for obstructive uropathy. Because the catheter is placed under ultrasound guidance, it is frequently malpositioned, and cases of fetal injury and death have been reported. 3,6There is also a high incidence of displacement and obstruction, requiring multiple reinsertions, thus increasing the risks of fetal injury and infection7 These disadvantages make catheter decompression useful only for short-term treatment late in gestation (ie, beyond 28 weeks’ gestation). We have devised a technique to overcome some of the risks of open fetal surgery and to specifically address the problem of fetal urinary tract obstruction. Transuterine endoscopy with a telescopic lens and operating instruments is used to create a vesicocutaneous fistula, through which is passed an expandable wire mesh cylinder. Placement under direct endoscopic vision insures optimal positioning for vesicoamniotic shunting and minimizes fetal injury. The endoscopic approach can also decrease the risks of fetal surgery by inducing less uterine trauma, thereby decreasing the likelihood of preterm labor and fetal morbidity from open surgery. Journal
of Pediatric
Surgery,
Vol27, No 8 (August), 1992: pp 950-954
FETOSCOPIC SURGERY FOR CONGENITAL
MATERIALS
ANOMALIES
AND METHODS
Two time-dated pregnant ewes (carrying a total of three male fetuses) at 75 days’ gestation (term = 145 days) were anesthetized with ketamine (1 g intramuscularly) and underwent general endotracheal anesthesia as previously described.8 Urethral and urachal obstruction were created 1 week prior to stent placement. The animals underwent laparotomy and hysterotomy and the fetal lower extremities were delivered through the uterine incision. A complete bladder obstruction was created by ligating the urethra with a 4-O silk ligature. The urachus was exposed at the inferior base of the umbilical cord, carefully dissected from the umbilical arteries, and also ligated with 4-O silk. The fetus was returned to the uterus. Lost amniotic fluid was replaced with warm Ringer’s lactate and 1 g of penicillin was added. The uterus was closed with a TA-55 stapling device (Ethicon, Inc, Cincinnati, OH), incorporating all layers, and returned to the abdomen, which was closed in layers. The ewe was awakened from anesthesia and returned to the stall where food and water were provided ad libitum. The endoscopic procedure was performed at a second operation 7 days later by exposing the uterus as described above. The gravid horn of the bicornuate uterus was palpated and the orientation of the fetus determined. At the proposed site of the camera port a 2-O silk purse-string suture was placed incorporating all layers of the uterus and membranes, and the suture threads passed through a rubber shod. A Veress needle was inserted perpendicularly into the uterus through all layers and into the amniotic space. A single distinct pop was noted and amniotic fluid was aspirated to confirm positioning. CO2 insufflation was begun at a SlOWrate and maintained at a pressure of no more than 3 to 5 cm HrO. After approximately 0.6 L of insufflation a satisfactory pneumometrium was obtained and the Veress needle was removed. A 5-mm trocar was then inserted through the purse-string, which was cinched down. A 5-mm 0” telescopic lens, connected to a xenon light source, was passed through this port and used to identify the optimal site for the accessory port, which was introduced in a similar manner (Fig 1). Positioning of the fetus was accomplished by external manipulation. All endoscopic equipment was provided by Karl Storz Endoscopy (Culver City, CA).
/
\
lnsufflated L_j_
retaining
\
Fig 2. The arrangement of the telescopic lens and manipulating instrument 8s seen through the uterus. lnsuffl8tion is wed to create a gss pocket t0 fllCilit8teviSU8k8tion. A camerr iS placed on the IenS eyepiece and the procedure is obsarved on a video monitor.
The urachal and urethral sutures were identified inferior to the root of the umbilical cord. A point was selected on the midline of the abdominal wall midway between the urethral meatus and the pubis, and the overlying skin and fascia were incised with unipolar cautery. A 16-gauge needle was passed through the uterus and introduced into the bladder under direct endoscopic vision. Once urine was aspirated, a .035-in diameter “J” wire was passed into the bladder and the needle removed. A Wallstent introducer (Schneider U.S. Stent, Inc, Minneapolis, MN) was passed over the guide wire into the bladder (Fig 2). The stent is 32 mm long and 8 mm wide and once deployed the wire mesh construction and flared ends securely maintain patency of the fistula. The stent was introduced under direct endoscopic vision, and urine was observed leaking out through the stem’s lumen (Fig 3). At the end of the procedure the uterus was deflated and 50 mL of warmed Ringer’s lactate was added containing 1 g of penicillin. The purse-string
Manipulating instrument
I
I
suture
/ \
‘.I , fl
-
Telescooi‘c lens [and yd&p camera
Fig 1. A l8parotomy is used to expose the gravid uterus. This illustrrtes 8 typical 8rrsngement of the camera port 8nd mllnipulating instrument on the anesthetized ewe.
Fig 3. The intrauterine view as seen through the endoscopa. The Wrllstent is seen being deployed into the vesicocut8neour ffstula. A, Wallstent; 6, umbilical cord; C, scrotum.
ESTES ET AL
952
sutures were tied and the uterine closure reinforced with interrupted 2-O silk sutures. The uterus was returned to the abdomen, which was closed as described above. RESULTS
Endoscopic visualization was initially attempted through amniotic fluid, and the results were unsatisfactory. The particulate debris in amniotic fluid produced excessive light scatter and poor visualization. Cautery was also ineffective due to the electrolyte-rich composition of amniotic fluid. These problems were solved by operating in a gas-filled space. There was no fetal or maternal morbidity. Urethral and urachal ligation were successful in creating a complete urinary obstruction in all three animals. Figure 4A demonstrates the dilated bladder and collecting system of the unstented, control fetus, 14 days after creation of the obstruction. Stents were placed endoscopically 7 days after urinary ligation in the experimental group (n = 2). Stent placement was confirmed intraoperatively by the return of urine from the dilated bladder. Figure 5 shows the stent passing into the decompressed blad-
Fig 5. En bloc removal of the bladder and overlying skin shows the stent passing through the abdominal wall into the bladder. A moderate amount of fibrosis was noted fixing the device in place. A, abdominal wall; B, bladder.
Fig 4. (A) Appearance of the urinary system in the unstented control fetus. The bladder, ureters, and renal pelves are moderately dilated. There was also an edenSiVe amOUnt Of retroperitoneal edema. (B) Appearance of urinary system in the stented fetus. The upper tracts and bladder are significantly less dilated than that of the control (Fig 4A). A, bladder; B, ureters; C, kidneys; D, right renal pelvis.
FETOSCOPIC SURGERY FOR CONGENITAL
ANOMALIES
der. The stem was patent and firmly seated in place by a fibrotic tissue reaction, and there was no evidence of peritoneal urine leakage. In both the experimental animals the urinary system appeared substantially less dilated (Fig 4B) than the control. DISCUSSION
This study demonstrates the feasibility of fetoscopic surgery. The endoscopic approach affords excellent visibility due to the coaxial light source and high magnification. For these reasons gas insufflation was used, as initial trials demonstrated excessive light scatter and distorted optics when visualizing through amniotic fluid. The air pocket creates a space in which surgical manipulation can easily be performed by displacing the uterine wall away from the fetus and allows for the effective use of cautery. There were several technical obstacles to overcome during the developmental phase of the procedure. Gaseous insufflation of the uterine space is potentially harmful to the fetus. In earlier studies we had several fetal deaths from excessive insufflation, probably from placenta-uterine separation. Because of this potential, we do not exceed 5 cm HZ0 during insufflation and excessive uterine distension is avoided. By following these guidelines we have had no subsequent fetal deaths. In addition, it is important to plan the surgical approach carefully. The fetus should be positioned prior to insufflation and held by external fixation if possible. The camera and instrument ports should be placed in a favorable geometry relative to the fetal region of interest, because the surgical field is severely constrained once the instruments are in place. An excellent example of the general problem of prenatal surgery is the need for improved management of fetuses with bilateral hydronephrosis. Fetal surgery has been successful in a few highly selected cases, but because preterm labor remains a serious obstacle, non-life-threatening malformations are not treated at present.1 Catheter drainage, on the other hand, is only useful for short-term therapy late in gestation because of the high incidence of obstruction and displacement.” In this study we successfully placed a bladder stent in midgestation fetal lambs with obstructive uropathy, with no fetal or maternal morbidity. The novel stenting device effectively maintained bladder drainage, without evidence of urinary leakage or displacement. The technique we have described for bladder drainage is less invasive than open fetal manipulation and potentially superior to catheter drainage because the stent is accurately placed under direct vision, and the large size and firm fixation make it unlikely to occlude or migrate during prolonged use.
953
Because this device can be used for long-term bladder drainage, we could apply the technique to the fetus with urinary obstruction earlier in pregnancy, and potentially salvage many fetuses who would eventually progress to irreversible renal and pulmonary dysfunction. The optical magnification provided by endoscopy facilitates procedures in early gestation fetuses, and we have demonstrated this by successfully creating a cleft lip model in a first trimester fetal lamb at 45 days’ gestation (unpublished results). Experimental studies demonstrate prevention of renal dysplasia and pulmonary hypoplasia by in utero decompression. These results also indicate that the degree of renal injury is proportional to the duration of the obstruction,4.9 thereby reinforcing the need for earlier intervention. We have developed several other fetoscopic procedures. A model of congenital cleft lip was created and repaired endoscopically in midgestation fetal lambs.“’ We have also placed indwelling transuterine catheters into placental vessels under endoscopic guidance and have created several models of congenital malformations including diaphragmatic hernia and gastroschisis in the first trimester. Fetoscopic surgery is performed through small uterine puncture sites, rather than a lo- to U-cm hysterotomy used presently for fetal exposure. By substantially reducing uterine trauma from the operation we expect to decrease perioperative uterine irritability and preterm labor. However, the safety of endoscopic surgery for the fetus has not been completely established. Work already in progress in our laboratory is focusing on potential fetal eye injury from the high intensity xenon light source. We are also examining the effects of CO2 and N20 insufflation on fetal hemodynamics and acid-base balance in a rhesus monkey model, which has a thicker and less compliant uterus than the sheep. Finally, techniques for percutaneous, transuterine port placement need to be developed so maternal laparotomy can be avoided completely. Endoscopic fetal surgery offers considerable promise for the future of antenatal therapy. By decreasing the risks of fetal manipulation, this technique may allow correction of non-life-threatening malformations in utero, such as cleft lip and palate, neural tube defects, and amniotic bands, as well as provide vascular access to the sick fetus afflicted with growth retardation or hydrops. In utero repair also offers the potential for scarless fetal wound healing, a remarkable property clinically unexploited.” Future research will undoubtedly expand the possibilities of fetoscopic surgery.
ESTES ET AL
954
REFERENCES 1. Harrison MR, Adzick NS: The fetus as a patient. Ann Surg 213:279-291,199l 2. Flake AW, Adzick NS, Glick PL, et al: Evaluation of the fetus with hydronephrosis, in deVere White RW, Palmer JM (eds): New Techniques in Urology. Mount Kisco, NY, Futura, 1987, pp 269-286 3. Harrison MR, Golbus MS, Filly RA, et al: Fetal hydronephrosis: Selection and surgical repair. J Pediatr Surg 22:556-558,1987 4. Crombleholme TM, Harrison MR, Longaker MT, et al: Prenatal diagnosis and management of bilateral hydronephrosis. Pediatr Nephrol2:334-342, 1988 5. Harrison MR, Nakayama DK, Noall R: Correction of congenital hydronephrosis in utero. II. Decompression reverses the effects of obstruction on the fetal lung and urinary tract. J Pediatr Surg 17:965-974,198l 6. Manning FA, Harrison MR, Rodeck C, et al: Catheter shunts
for fetal hydronephrosis and hydrocephalus. Report of the International Fetal Surgery Registry. N Engl J Med 315:336-340,1986 7. Glick PL, Harrison MR, Golbus MS, et al: Management of the fetus with congenital hydronephrosis II: Prognostic criteria and selection for treatment. J Pediatr Surg 20~376-387,1985 8. Longaker MT, Whitby DJ, Adzick NS, et al: Studies in fetal wound healing, VI. Second and early third trimester fetal wounds demonstrate rapid collagen deposition without scar formation. J Pediatr Surg 25:63-69,199O 9. Glick PL, Harrison MR, Adzick NS, et al: Correction of congenital hydronephrosis in utero IV: In utero decompression prevents renal dysplasia. J Pediatr Surg 19:649-657, 1984 10. Estes JM, Whitby DJ, Lorenz HP, et al: Endoscopic creation and repair of fetal cleft lip. Plast Reconstr Surg (in press) 11. Longaker MT, Adzick NS: The biology of fetal wound healing: A review. Plast Reconstr Surg 87:788-798, 1991
Discussion Steve Golladay (New Orleans, LA): I find their work is seen as often now in Newsweek and Time as in other medical journals. They have combined the technique of fetal surgery and cephalocentesis first described by Og of Central Africa in Paleolithic times with laparoscopy from Kelling of Dresden in 1901 to produce fetoscopy. Open fetal procedures were performed in three male lambs at 75 days gestation with fetoscopic procedures 7 days later. The scope part of the procedure is intended to decrease fetal risks and reduce preterm labor. The sheep is easily operable above about 60 days’ gestation without causing fatal surgery instead of fetal surgery; at 82 days they did the fetoscopic procedures, approximately the equivalent of 24 to 25 weeks’ human gestation, a time at which successful human procedures have been done. And following uneventful open procedures, the fetoscopy was accomplished at a time which we would expect no problems from open procedures. Have earlier gestational and riskier procedures been done? The lambs were only two, making P value for risk very difficult to document. Do we have evidence of decreased risk to the mother or fetus? The stent was said to stay in place better and last longer, but how is this better and for how long? That was not found in the manuscript. The technique is described elegantly and brilliantly detailed in the manuscript, and it may prove the advent of the future from a technical standpoint. However, with increasing government regulations for some forms of fetal surgery such as abortion, what do you consider to be the future of fetal surgery today? James ht. Estes (response): I will first address the
question regarding early fetal intervention. In our experience open procedures are nearly 100% fatal in lambs prior to 60 days’ gestation, probably due to their extreme susceptibility and sensitivity to hypothermia. Due to its less invasive nature, we believe endoscopic fetal surgery will allow intervention earlier in gestation. In fact, I have performed an endoscopic procedure on two fetal lambs at 45 days’ gestation (equivalent to the first trimester) with no perioperative morbidity. Clearly this question merits further investigation. Our study used a small number of animals to demonstrate the feasibility and efficacy of a new technique. By no means did we attempt to recapitulate previous work which has already established the natural history and rationale for in utero correction of obstructive uropathy. Furthermore, we have subsequently gained additional experience with this technique that confirms our initial observations. We left the stent in place 7 days prior to examination. At that time it was widely patent and firmly fixed in place by tissue growth into the interstices of the wire mesh. The stent’s large diameter is an inherent advantage over the standard 4F double-pigtail catheter, which commonly fails due to obstruction. This leads us to believe that the stent would provide more reliable bladder drainage. The efficacy of fetal surgery is well established in animal models for congenital diaphragmatic hernia (CDH) and bladder obstruction. The future of in utero surgery in humans awaits a more critical evaluation and we are presently conducting a randomized, prospective trial comparing fetal surgery to conventional management for CDH.
Surgery for the Treatment
of Congenital
Anomalies
By James M. Estes, Thomas E. MacGillivray, Marc H. Hedrick, N. Scott Adzick, and Michael R. Harrison San Francisco, California l Fetoscopic techniques may broaden the indications for prenatal surgical intervention by obviating the risks of hysterotomy. For example, congenital obstructive uropathy has been treated by open vesicostomy and percutaneous catheter placement. The open approach is appropriate only for highly selected fetuses because of the inherent risks, whereas catheter drainage, though a safer procedure, is only useful for short-term therapy late in gestation due to frequent catheter obstruction and migration. The natural history of congenital obstructive uropathy mandates the need for improved therapy earlier in gestation, in order to salvage fetuses who would otherwise die of renal failure and pulmonary hypoplasia. We have developed a potential solution to this problem in which surgery is performed on the fetus without the risks of hysterotomy. Endoscopic fetal surgery uses a telescopic lens and operating instruments that are passed through small “ports“ in the uterus. A bubble of CO2 is used to displace amniotic fluid and provides excellent visualization in a magnified field. This approach is considerably less invasive than open fetal surgery and, therefore, is less likely to provoke preterm labor. In this study we corrected obstructive uropathy in midgestation fetal lambs using a new, expandable wire mesh stent that is placed endoscopically and should provide more reliable bladder drainage than existing catheters. The fetoscopic surgicaf approach can potentially expand the indications for in utero surgery by decreasing fetal risks, facilitating intervention earlier in gestation, and reducing preterm labor. As a consequence, the potential now exists to correct non-life-threatening malformations in utero. Copyright o 1992 by W. B. Saunders Company INDEX WORDS:
Fetoscopy;
obstructive
uropathy;
fetal sur-
gery.
I
NVASIVE techniques for prenatal fetal therapy are relatively limited. Open fetal surgery is an option for the highly selected fetus with a lifethreatening malformation, but carries significant morbidity predominantly from preterm labor.’ On the opposite end of the therapeutic spectrum are percutaneous techniques, which are used to drain fluid collections and place catheters. Failures due to tech-
From the Depanrnent of Surgery and the Fetal Treatment !+ograrn, University of California, San Francisco, San Francisco, CA. Third Place Winner in the Jens G. Rosencraru Resident Competition. Presented at the 43rd Annual Meeting of the Surgical Section of the Ametican Academy of Pediatrics, New Orleans, Louisiana, October 26-27, 1991. Address reprint requests to James M. Estes, MD, Fetal Treatment Program, Room HSE-585, University of California, 513 Pamassus St, San Francisco, CA 94143-0570. Copyright o 1992 by WE. Saunders Company 0022-3468/92/2708-0003$03.00/0
950
nical problems such as catheter obstruction and migration are common, limiting this procedure only to short-term therapy. The current management of congenital obstructive uropathy illustrates these problems well. It is a common malformation affecting approximately one of every one thousand fetuses, and causes death in the most severely affected.2 Fetal surgery has been successful with a few highly selected cases of this anomaly. The alternate treatments, which include percutaneous bladder stenting and repeated aspiration, offer only short-term palliation due to the high incidence of failure.3 Hydronephrosis is easily diagnosed in utero. Unilateral hydronephrosis is managed conservatively, while bilateral disease can present as a continuum of severity. Criteria have been developed to assess renal function in utero and these parameters include urine electrolytes, ultrasonographic appearance of kidneys (presence or absence of dysplastic cysts), urine output, and amniotic fluid volume.4 Several case reports in humans and experimental work in lambs demonstrate that in utero bladder drainage may restore amniotic fluid volume and prevent progressive renal deterioration and pulmonary hypoplasia,4J thus providing the scientific rationale for our study. Percutaneous transabdominal placement of a catheter is the most common procedure performed in utero for obstructive uropathy. Because the catheter is placed under ultrasound guidance, it is frequently malpositioned, and cases of fetal injury and death have been reported. 3,6There is also a high incidence of displacement and obstruction, requiring multiple reinsertions, thus increasing the risks of fetal injury and infection7 These disadvantages make catheter decompression useful only for short-term treatment late in gestation (ie, beyond 28 weeks’ gestation). We have devised a technique to overcome some of the risks of open fetal surgery and to specifically address the problem of fetal urinary tract obstruction. Transuterine endoscopy with a telescopic lens and operating instruments is used to create a vesicocutaneous fistula, through which is passed an expandable wire mesh cylinder. Placement under direct endoscopic vision insures optimal positioning for vesicoamniotic shunting and minimizes fetal injury. The endoscopic approach can also decrease the risks of fetal surgery by inducing less uterine trauma, thereby decreasing the likelihood of preterm labor and fetal morbidity from open surgery. Journal
of Pediatric
Surgery,
Vol27, No 8 (August), 1992: pp 950-954
FETOSCOPIC SURGERY FOR CONGENITAL
MATERIALS
ANOMALIES
AND METHODS
Two time-dated pregnant ewes (carrying a total of three male fetuses) at 75 days’ gestation (term = 145 days) were anesthetized with ketamine (1 g intramuscularly) and underwent general endotracheal anesthesia as previously described.8 Urethral and urachal obstruction were created 1 week prior to stent placement. The animals underwent laparotomy and hysterotomy and the fetal lower extremities were delivered through the uterine incision. A complete bladder obstruction was created by ligating the urethra with a 4-O silk ligature. The urachus was exposed at the inferior base of the umbilical cord, carefully dissected from the umbilical arteries, and also ligated with 4-O silk. The fetus was returned to the uterus. Lost amniotic fluid was replaced with warm Ringer’s lactate and 1 g of penicillin was added. The uterus was closed with a TA-55 stapling device (Ethicon, Inc, Cincinnati, OH), incorporating all layers, and returned to the abdomen, which was closed in layers. The ewe was awakened from anesthesia and returned to the stall where food and water were provided ad libitum. The endoscopic procedure was performed at a second operation 7 days later by exposing the uterus as described above. The gravid horn of the bicornuate uterus was palpated and the orientation of the fetus determined. At the proposed site of the camera port a 2-O silk purse-string suture was placed incorporating all layers of the uterus and membranes, and the suture threads passed through a rubber shod. A Veress needle was inserted perpendicularly into the uterus through all layers and into the amniotic space. A single distinct pop was noted and amniotic fluid was aspirated to confirm positioning. CO2 insufflation was begun at a SlOWrate and maintained at a pressure of no more than 3 to 5 cm HrO. After approximately 0.6 L of insufflation a satisfactory pneumometrium was obtained and the Veress needle was removed. A 5-mm trocar was then inserted through the purse-string, which was cinched down. A 5-mm 0” telescopic lens, connected to a xenon light source, was passed through this port and used to identify the optimal site for the accessory port, which was introduced in a similar manner (Fig 1). Positioning of the fetus was accomplished by external manipulation. All endoscopic equipment was provided by Karl Storz Endoscopy (Culver City, CA).
/
\
lnsufflated L_j_
retaining
\
Fig 2. The arrangement of the telescopic lens and manipulating instrument 8s seen through the uterus. lnsuffl8tion is wed to create a gss pocket t0 fllCilit8teviSU8k8tion. A camerr iS placed on the IenS eyepiece and the procedure is obsarved on a video monitor.
The urachal and urethral sutures were identified inferior to the root of the umbilical cord. A point was selected on the midline of the abdominal wall midway between the urethral meatus and the pubis, and the overlying skin and fascia were incised with unipolar cautery. A 16-gauge needle was passed through the uterus and introduced into the bladder under direct endoscopic vision. Once urine was aspirated, a .035-in diameter “J” wire was passed into the bladder and the needle removed. A Wallstent introducer (Schneider U.S. Stent, Inc, Minneapolis, MN) was passed over the guide wire into the bladder (Fig 2). The stent is 32 mm long and 8 mm wide and once deployed the wire mesh construction and flared ends securely maintain patency of the fistula. The stent was introduced under direct endoscopic vision, and urine was observed leaking out through the stem’s lumen (Fig 3). At the end of the procedure the uterus was deflated and 50 mL of warmed Ringer’s lactate was added containing 1 g of penicillin. The purse-string
Manipulating instrument
I
I
suture
/ \
‘.I , fl
-
Telescooi‘c lens [and yd&p camera
Fig 1. A l8parotomy is used to expose the gravid uterus. This illustrrtes 8 typical 8rrsngement of the camera port 8nd mllnipulating instrument on the anesthetized ewe.
Fig 3. The intrauterine view as seen through the endoscopa. The Wrllstent is seen being deployed into the vesicocut8neour ffstula. A, Wallstent; 6, umbilical cord; C, scrotum.
ESTES ET AL
952
sutures were tied and the uterine closure reinforced with interrupted 2-O silk sutures. The uterus was returned to the abdomen, which was closed as described above. RESULTS
Endoscopic visualization was initially attempted through amniotic fluid, and the results were unsatisfactory. The particulate debris in amniotic fluid produced excessive light scatter and poor visualization. Cautery was also ineffective due to the electrolyte-rich composition of amniotic fluid. These problems were solved by operating in a gas-filled space. There was no fetal or maternal morbidity. Urethral and urachal ligation were successful in creating a complete urinary obstruction in all three animals. Figure 4A demonstrates the dilated bladder and collecting system of the unstented, control fetus, 14 days after creation of the obstruction. Stents were placed endoscopically 7 days after urinary ligation in the experimental group (n = 2). Stent placement was confirmed intraoperatively by the return of urine from the dilated bladder. Figure 5 shows the stent passing into the decompressed blad-
Fig 5. En bloc removal of the bladder and overlying skin shows the stent passing through the abdominal wall into the bladder. A moderate amount of fibrosis was noted fixing the device in place. A, abdominal wall; B, bladder.
Fig 4. (A) Appearance of the urinary system in the unstented control fetus. The bladder, ureters, and renal pelves are moderately dilated. There was also an edenSiVe amOUnt Of retroperitoneal edema. (B) Appearance of urinary system in the stented fetus. The upper tracts and bladder are significantly less dilated than that of the control (Fig 4A). A, bladder; B, ureters; C, kidneys; D, right renal pelvis.
FETOSCOPIC SURGERY FOR CONGENITAL
ANOMALIES
der. The stem was patent and firmly seated in place by a fibrotic tissue reaction, and there was no evidence of peritoneal urine leakage. In both the experimental animals the urinary system appeared substantially less dilated (Fig 4B) than the control. DISCUSSION
This study demonstrates the feasibility of fetoscopic surgery. The endoscopic approach affords excellent visibility due to the coaxial light source and high magnification. For these reasons gas insufflation was used, as initial trials demonstrated excessive light scatter and distorted optics when visualizing through amniotic fluid. The air pocket creates a space in which surgical manipulation can easily be performed by displacing the uterine wall away from the fetus and allows for the effective use of cautery. There were several technical obstacles to overcome during the developmental phase of the procedure. Gaseous insufflation of the uterine space is potentially harmful to the fetus. In earlier studies we had several fetal deaths from excessive insufflation, probably from placenta-uterine separation. Because of this potential, we do not exceed 5 cm HZ0 during insufflation and excessive uterine distension is avoided. By following these guidelines we have had no subsequent fetal deaths. In addition, it is important to plan the surgical approach carefully. The fetus should be positioned prior to insufflation and held by external fixation if possible. The camera and instrument ports should be placed in a favorable geometry relative to the fetal region of interest, because the surgical field is severely constrained once the instruments are in place. An excellent example of the general problem of prenatal surgery is the need for improved management of fetuses with bilateral hydronephrosis. Fetal surgery has been successful in a few highly selected cases, but because preterm labor remains a serious obstacle, non-life-threatening malformations are not treated at present.1 Catheter drainage, on the other hand, is only useful for short-term therapy late in gestation because of the high incidence of obstruction and displacement.” In this study we successfully placed a bladder stent in midgestation fetal lambs with obstructive uropathy, with no fetal or maternal morbidity. The novel stenting device effectively maintained bladder drainage, without evidence of urinary leakage or displacement. The technique we have described for bladder drainage is less invasive than open fetal manipulation and potentially superior to catheter drainage because the stent is accurately placed under direct vision, and the large size and firm fixation make it unlikely to occlude or migrate during prolonged use.
953
Because this device can be used for long-term bladder drainage, we could apply the technique to the fetus with urinary obstruction earlier in pregnancy, and potentially salvage many fetuses who would eventually progress to irreversible renal and pulmonary dysfunction. The optical magnification provided by endoscopy facilitates procedures in early gestation fetuses, and we have demonstrated this by successfully creating a cleft lip model in a first trimester fetal lamb at 45 days’ gestation (unpublished results). Experimental studies demonstrate prevention of renal dysplasia and pulmonary hypoplasia by in utero decompression. These results also indicate that the degree of renal injury is proportional to the duration of the obstruction,4.9 thereby reinforcing the need for earlier intervention. We have developed several other fetoscopic procedures. A model of congenital cleft lip was created and repaired endoscopically in midgestation fetal lambs.“’ We have also placed indwelling transuterine catheters into placental vessels under endoscopic guidance and have created several models of congenital malformations including diaphragmatic hernia and gastroschisis in the first trimester. Fetoscopic surgery is performed through small uterine puncture sites, rather than a lo- to U-cm hysterotomy used presently for fetal exposure. By substantially reducing uterine trauma from the operation we expect to decrease perioperative uterine irritability and preterm labor. However, the safety of endoscopic surgery for the fetus has not been completely established. Work already in progress in our laboratory is focusing on potential fetal eye injury from the high intensity xenon light source. We are also examining the effects of CO2 and N20 insufflation on fetal hemodynamics and acid-base balance in a rhesus monkey model, which has a thicker and less compliant uterus than the sheep. Finally, techniques for percutaneous, transuterine port placement need to be developed so maternal laparotomy can be avoided completely. Endoscopic fetal surgery offers considerable promise for the future of antenatal therapy. By decreasing the risks of fetal manipulation, this technique may allow correction of non-life-threatening malformations in utero, such as cleft lip and palate, neural tube defects, and amniotic bands, as well as provide vascular access to the sick fetus afflicted with growth retardation or hydrops. In utero repair also offers the potential for scarless fetal wound healing, a remarkable property clinically unexploited.” Future research will undoubtedly expand the possibilities of fetoscopic surgery.
ESTES ET AL
954
REFERENCES 1. Harrison MR, Adzick NS: The fetus as a patient. Ann Surg 213:279-291,199l 2. Flake AW, Adzick NS, Glick PL, et al: Evaluation of the fetus with hydronephrosis, in deVere White RW, Palmer JM (eds): New Techniques in Urology. Mount Kisco, NY, Futura, 1987, pp 269-286 3. Harrison MR, Golbus MS, Filly RA, et al: Fetal hydronephrosis: Selection and surgical repair. J Pediatr Surg 22:556-558,1987 4. Crombleholme TM, Harrison MR, Longaker MT, et al: Prenatal diagnosis and management of bilateral hydronephrosis. Pediatr Nephrol2:334-342, 1988 5. Harrison MR, Nakayama DK, Noall R: Correction of congenital hydronephrosis in utero. II. Decompression reverses the effects of obstruction on the fetal lung and urinary tract. J Pediatr Surg 17:965-974,198l 6. Manning FA, Harrison MR, Rodeck C, et al: Catheter shunts
for fetal hydronephrosis and hydrocephalus. Report of the International Fetal Surgery Registry. N Engl J Med 315:336-340,1986 7. Glick PL, Harrison MR, Golbus MS, et al: Management of the fetus with congenital hydronephrosis II: Prognostic criteria and selection for treatment. J Pediatr Surg 20~376-387,1985 8. Longaker MT, Whitby DJ, Adzick NS, et al: Studies in fetal wound healing, VI. Second and early third trimester fetal wounds demonstrate rapid collagen deposition without scar formation. J Pediatr Surg 25:63-69,199O 9. Glick PL, Harrison MR, Adzick NS, et al: Correction of congenital hydronephrosis in utero IV: In utero decompression prevents renal dysplasia. J Pediatr Surg 19:649-657, 1984 10. Estes JM, Whitby DJ, Lorenz HP, et al: Endoscopic creation and repair of fetal cleft lip. Plast Reconstr Surg (in press) 11. Longaker MT, Adzick NS: The biology of fetal wound healing: A review. Plast Reconstr Surg 87:788-798, 1991
Discussion Steve Golladay (New Orleans, LA): I find their work is seen as often now in Newsweek and Time as in other medical journals. They have combined the technique of fetal surgery and cephalocentesis first described by Og of Central Africa in Paleolithic times with laparoscopy from Kelling of Dresden in 1901 to produce fetoscopy. Open fetal procedures were performed in three male lambs at 75 days gestation with fetoscopic procedures 7 days later. The scope part of the procedure is intended to decrease fetal risks and reduce preterm labor. The sheep is easily operable above about 60 days’ gestation without causing fatal surgery instead of fetal surgery; at 82 days they did the fetoscopic procedures, approximately the equivalent of 24 to 25 weeks’ human gestation, a time at which successful human procedures have been done. And following uneventful open procedures, the fetoscopy was accomplished at a time which we would expect no problems from open procedures. Have earlier gestational and riskier procedures been done? The lambs were only two, making P value for risk very difficult to document. Do we have evidence of decreased risk to the mother or fetus? The stent was said to stay in place better and last longer, but how is this better and for how long? That was not found in the manuscript. The technique is described elegantly and brilliantly detailed in the manuscript, and it may prove the advent of the future from a technical standpoint. However, with increasing government regulations for some forms of fetal surgery such as abortion, what do you consider to be the future of fetal surgery today? James ht. Estes (response): I will first address the
question regarding early fetal intervention. In our experience open procedures are nearly 100% fatal in lambs prior to 60 days’ gestation, probably due to their extreme susceptibility and sensitivity to hypothermia. Due to its less invasive nature, we believe endoscopic fetal surgery will allow intervention earlier in gestation. In fact, I have performed an endoscopic procedure on two fetal lambs at 45 days’ gestation (equivalent to the first trimester) with no perioperative morbidity. Clearly this question merits further investigation. Our study used a small number of animals to demonstrate the feasibility and efficacy of a new technique. By no means did we attempt to recapitulate previous work which has already established the natural history and rationale for in utero correction of obstructive uropathy. Furthermore, we have subsequently gained additional experience with this technique that confirms our initial observations. We left the stent in place 7 days prior to examination. At that time it was widely patent and firmly fixed in place by tissue growth into the interstices of the wire mesh. The stent’s large diameter is an inherent advantage over the standard 4F double-pigtail catheter, which commonly fails due to obstruction. This leads us to believe that the stent would provide more reliable bladder drainage. The efficacy of fetal surgery is well established in animal models for congenital diaphragmatic hernia (CDH) and bladder obstruction. The future of in utero surgery in humans awaits a more critical evaluation and we are presently conducting a randomized, prospective trial comparing fetal surgery to conventional management for CDH.