- Email: [email protected]
Early experience with open fetal surgery for congenital hydronephrosis
Early Experience With Open Fetal Surgery for Congenital Hydronephrosis By Timothy M. Crombleholme, Michael R. Harrison, Jacob C. Langer, Michael T. Longaker, Robert L. Anderson, Nathan S. Slotnick, Roy A. Filly, Peter W. Callen, Ruth B. Goldstein, and Mitchell S. Golbus San Francisco, California
9 The fetus w i t h severe bilateral hydronephrosis and 9 a s s o c i a t e d oligohydramnios in t h e s e c o n d t r i m e s t e r i s . d o o m e d a t b i r t h by ongoing p u l m o n a r y a n d renal d a m a g e . Since decompression w i t h percutaneously placed catheters is not dependable for long periods, w e have developed anesthetic, surgical, and tocolytic techniques for open fetal 9 urinary tract decompression in animals, and have n o w applied those techniques to a small group of five patients.. One had bilateral ureterostomies and the subsequent f o u r , had marsupialization of the bladder. All pregnancies proceeded to cesarean delivery at 32 to 35 weeks" gestation. There was no long-term maternal morbidity, and t w o mothers have since experienced normal pregnancies. Three fetuses had return of normal amniotic fluid dynamics, and all three had adequate pulmonary function at birth, suggesting~that fatal pulmonary hypoplasia associated w i t h early severe oligohydramnios had boon reversed. ~ w o neonates died at birth w i t h pulmonary~ hypoplasia. One had no amniotic fluid even after decompression, and the other had some amniotic fluid after decompression but a tiny chest cavity due to the long period of severe oligohydramnios before decompression. Of the three surviving infants, one had normal renal func-. tion when she died of unrelated causes at 9 months of age. One has normal renal function at 23 months and the third had failing renal function at 21/2 years and has grown and developed normally, but will require renal transplantation. We have n o w developed selection criteria that would exclude from t r e a t m e n t the t w o fetuses w h o died of pulmonary hypoplasia and the one w h o developed renal 9 failure. Our initial limited experience suggests the following: (1) that open fetal surgery is safe in experienced hands and the risks of open fetal surgery can be minimized by extensive prepatory experimental w o r k ; (2) that in a f e w highly selected cases, open decompression can restore amniotic fluid dynamics and is efficacious in preventing pulmonary hypoplasia at birth; and (3)' that the effect of decompression on the development of renal dysplasia and. ult.mate renal function remahs unknown. 9 1988 b y (;rune & Stratton, inc.
been established by many experimental studies? "6 These studies demonstrate that potentially fatal pulmonary hypoplasia, caused by oligohydramnios resulting from obstructive uropathy, may be ameliorated by decompression in utero. 1'2'5'6 In addition, in-utero decompression may arrest dysplastic morphogenesis of the kidney and preserve renal function depending on the timing and the duration of the obstruction) "4 The efficacy of this procedure in human fetuses is yet to be established9 Vesicoamniotic shunts have been employed in the short-term treatment of bilateral fetal hydronephrosis, but due to the high incidence of catheter clogging or displacement and the associated risk of chorioamnionitis, vesicoamniotic shunts are not satisfactory for longterm fetal urinary tract decompression. 7"~2 Severely affected fetuses with associated oligohydramnios diagnosed prior to 24 weeks' gestation generally have not survived even when successfully decompressed. 9"~7Due to improvements in prognostic criteria, we may now be able to select for intervention those fetuses whose kidneys are not already irreversibly damaged and could, if decompressed, produce sufficient urine to restore amniotic fluid dynamics, preventing neonatal death from pulmonary hypoplasia. This brief review is a report of our initial experience with open fetal surgery performed in five cases of bilateral hydronephrosis. These cases illustrate that open fetal surgery is feasible with acceptable maternal risk, and that restoration of normal amniotic fluid dynamics can reverse potentially fatal pulmonary hypoplasia. Efficacy in reversing renal dysplasia is not known. MATERIALS AND METHODS
INDEX WORDS: Fetal surgery; congenital hydronephrosis.
T
HE THEORETICAL BASIS for in-utero decompression of bilateral fetal hydronephrosis has
From the Fetal Treatment Program and the Departments of Surgery, Obstetrics, Gynecology and Reproductive Science, and Radiology, University of California, San Francisco. Presented at the 19th Annual Meeting of the American Pediatric Surgical Association, Tucson, Arizona, May I 1-14, 1988. Address reprint requests to Michael R. Harrison. MD. The Fetal Treatment Program, 585 HSE, UCSF, Third and Parnassas Ave, San Francisco, CA 94143. 9 1988 by Grune & Stratton, Inc. 0022-3468/88/2312-0005503.00/0 1114
More than 200 cases of bilateral fetal hydronephrosis have been referred to the Fetal Treatment Program at the University of California, San Francisco Medical Center from 1978 to 1988. Five patients were found suitable for open intervention. Two of these cases have been previously reported. ]6']7All surgical procedures were performed by M.R.H.
CASE REPORTS
Case 1 A primagravida woman was referred with a male fetus (G.T.) of 20 weeks' gestation who had oligohydramnios and bilateral congenital hydronephrosis. Fetal bilateral ureterostomies were created, but oligohydranmios persisted and the neonate died of pulmonary hypoplasia. '~
Journal of Pediatric Surgery, Vo123, No 12 (December),1988: pp 1114-1121
OPEN FETAL SURGERY FOR CONGENITAL HYDRONEPHROSIS
1115
Case 2 A 31-year-old primagravida woman was referred from Texas with a male fetus of 22 weeks' gestation. The fetus (M.M.) had oligohydramnios and bilateral hydronephrosis. Open fetal surgery was successfully performed at 24 weeks' gestation with decompression of the urinary tract and restoration of amniotic fluid volume for the remainder of the gestation. The infant had no significant respiratory difficulty and has developed normally; however, renal function has deteriorated gradually. Although dialysis has not been necessary (the boy is now 2V2years old), transplantation is planned) 7
B
Pulse imeter
Case 3 A 19-year-old primagravida woman was referred from Minnesota at 22 weeks' gestation. The female fetus (M.S.) bad gross megacystis with severe bilateral hydronephrosis and slightly ecbogenic renal parenchyma. The patient's amniotic fluid volume at presentation was low and then disappeared entirely. Percutaneous fetal bladder drainage for four hours showed good urine production (4 mL/h) with a sodium level of 79 mEq/L, a chloride level of 71 mEq/L, and osmolarity of 150 mosm. This electrolyte pattern suggested good potenttal for functional renal recovery, l~ Karyotype was normal, 46XX. Over five days, the oligohydramnios continued to worsen and fetal bladder distention increased. The family was counseled about the options available including termination, observation, vesicoamniotic catheter shunt, and open fetal surgery. After extensive discussion of the risks of hysterotomy and fetal bladder marsupialization, and the experimental nature of the procedure, the family chose to undergo open decompression. The surgery was performed at 23 weeks' gestation using anesthetic and tocolytic regimens previuosly described. 16.17Briefly, this included indomethacin, haiothane for uterine relaxation, fetal and maternal anesthesia intraoperatively, and ritodrine infusion and/or indomethacin postoperatively. The uterus was exposed by a low transverse abdominal incision, and intraoperative sonography confirmed the fetal position as breech facing anterior and the placenta was posterior. A low transverse hysterotomy was made directly over the massively distended fetal abdomen and bladder, and due to the complete absence of amniotic fluid, the fetal abdomen was entered immediately. The bysterotomy was extended with a newly developed resorptive stapling device and the lower half of the fetus exterionized. The massively distended bladder was opened and sutured to the thinned out redundant abdominal wall with 3-0 silk, thus scaling the peritoneal cavity and marsupializing the bladder (Fig 1). The fetus was returned, with some difficulty due to the small amniotic cavity, and the uterus was dosed with all layer interrupted Dexan over a running Dexon inner layer. Amniotic fluid volume was restored with warm Ringer's lactate containing oxacillin. Fetal heart rate monitored sonographically remained remarkably stable throughout the procedure. The postoperative course was uneventful. Ritodrine was given intravenously for three days and then orally. The patient was discharged after nine days and returned to Minnesota. The fetal condition improved dramatically as serial ultrasound documented complete decompression of the fetal megacystis, hydronephrosis with minimal residual caliectasis, and normal amniotie fluid volume. The thin redundant abdominal wall ("prune belly") improved and interval growth was normal. Twenty-four days postoperatively the mother presented with abdominal pain peritoneal signs, and preterm labor. At laparotomy, her obstetrician found a small amniotic fluid leak from the hysterotomy incision which was easily controlled with a single O chromic suture. She was again discharged on oral tocolytics. At 31 weeks' gestation, she was readmitted with spontaneous premature rupture of membranes and maintained on oral ritodrine with bed rest, without evidence of
C
Fig 1. Example of open fetal surgery for hydronephrosis. IA) Bilateral hydronephrosis, megscystis, and oligohydramnios at approximately 24 weeks" gestation. Guided by intrsoperative sonography, a hysterotomy is placed to avoid the placenta and expose the fetus" lower body without disturbing the umbilical circulation. (B) Fetal heart rate and oxygen saturation are monitored by transcutaneous pulse oximeter and ECG while a cutaneous vesicostomy is constructed. (C) Amniotic fluid volume restored by fetal urine. 17
chorioanmionitis for 2 weeks. A cesarean section was performed at 33 weeks' gestation; the previous hysterotomy incision was found to be thinned but intact. The 4 lb 2 oz baby girl required minimal ventilatory support. Although her respiratory function was normal after birth, she required neonatal exploration and ileal diversion for mieroileum and microcolon. Her renal function remained normal, but persistent gastrointestinal malfunction suggested pseudoobstruction, which did not improve despite further surgical intervention and eventually led to death from septicemia at 9 months of age.
Case 4 This 35-year-old prirnagravida woman was found at 16 weeks' gestation to have oligohydramnios, and the female fetus (CH) had megacystis and bilateral hydronepbrosis. Ultrasound showed no evidence of cystic renal parenchymal changes. Aspirated fetal urine sodium concentration of 100 mEq/L and chloride concentration of 91 mEq/L suggested good potential for renal recovery, although osmolarity was slightly elevated at 221 mosm. Percutaneous umbilical blood sampling (PUBS) for karyotype analysis yielded a normal 46XX result. The family who had battled infertility for years, was counseled at length about all the available options. They resolutely refused termination and insisted on pursuing treatment, although we were reluctant to offer treatment to a female fetus since the pathophysioiogy was not clearly defined. At 18 weeks' gestation, open fetal surgery was performed as described in the previous cases. Intraoperative sonography showed the fetus in a transverse position wffh the lower extremities to the : left. Unfortunately, a placentae previa covered the entire anterior
1 1 16
CROMBLEHOLME ET AL
o 8
.o Z
z=
"~2
;I
o
r~
E~
E~
E~ c
iZ
| -~
0
!
~
o o
,~
o o
~
-
o
o
~
~
E
~-~
~ E LUre E E~
~ o'~ "e~ o ~ E ~ ~ E ~ 0 E~ Emm = e9 E~O E~
z~o
z~oz~oz~o
~.~
~
|
o
co
0
"o
,==.~.-=
~
o
~"
"o
|
|
>-
|
~
0
o
0
=~._~
|
~'6
>'-
=-
|
>
=
-
~ ~>..;
6 R "o
.>
OPEN FETAL SURGERY FOR CONGENITAL HYDRONEPHROSIS
wall of the uterus from cervix to fundus. The uterus was opened at the only placenta-free area on the left posterior wall of the fundus. The lower extremities of the fetus were exteriorized, and a transcutaneous pulse oximeter was placed around the fetal thigh. A midline suprapubie incision was made through the fetal abdomen, exposing a distended thick walled bladder.The bladder was opened and rnarsupializedto the fetalabdominal wall with interrupted4-0 silksutures. The fetus was returned to the uterine cavity within ten minutes. Amniotic fluid was restored with warm normal saline.The uterus was closed with interruptedO - P D S all layer sutures. After uterine closure, sonography showed a strong fetal heart rate of 140. Ritodrine hydrochloride infusionwas weaned within three days, and the patient was switched to oral terbutoline for control of uterine irritability.Both the mother and fetus recovered uneventfully.The patientleftthe hospitalaftereightdays, and returned to her home in southern California. The fetus grew normally. The bladder remained decompressed, and was normal upon subsequent ultrasound examinations. Ten
weeks after fetal surgery, at 28 weeks' gestation, the patient presented to her local medical center with vaginal bleeding and preterm labor thought to be due to mild abruptio placentae. The bleeding stopped with conservative measures, and preterm labor was controlled for the next 19 days with bed rest, ritodrine, and occasionally magnesium sulfate. The membranes ruptured at 32 weeks' gestation and a cesarean section was performed. The uterine incision was well healed. A healthy 1,750-g baby girl had Apgar scores of 9 at one minute and 9 at five minutes. Ultrasound showed a distended bladder and hydronephrosis but normal renal parenehyma. Neonatal creatinine was 1.1 mg/dL, but at 1 month fell to and remained at 0.6 mg/dL. Voiding cystourethrogram showed a hypertrophied bladder with dynamic evidence of neurogenic dysfunction and an elongated nonobstructed urethra. The infant's course has been uncomplicated except for a urinary tract infection at 3 weeks of age. She has continued to do well with normal renal function, and is now 2 years old.
Case 5 A 27-year-old gravida 4 para 3 mother was referred from New Mexico at 20 weeks' gestation with oligohydramnios, and a small (for gestational age) fetus (K.G.) with massive fetal abdominal distention, megaeystis, and bilateral hydronephrosis. Karyotype analysis yielded a normal (46XY) result. Both kidneys evidenced increased eehogenicity but no cortical cysts, and aspirated fetal urine had borderline values (Na, 100 mEq/L; CI, 82 mEq/L; osm, 215 mosm/kg). The family was counseled about the options available and the likelihood of a poor outcome regardless of the course chosen. The family adamantly refused termination and requested open fetal surgery despite the risks of an experimental procedure and the guarded prospects for a successful outcome. At 22 weeks' gestation, open fetal surgery was performed using the regimen described above. Intraoperative sonography showed the fetus was lying head down, with back anterior. Since the placenta was anterior near the top of the fundus, the hysterotomy was made posteriorly. The lower extremities were delivered into the wound and a transcutaneous pulse oximeter was applied to the fetal thigh. The vesicostomy was performed as in cases 2 to 4, and required 12 minutes of extrauterine exposure. The hysterotomy was closed with through-and-through O-PDS mattress sutures, and two layers of running Dexon. The amniotic fluid volume was restored with warm normal saline. Postoperative tocolysis included indomethaein per rectum, magnesium sulfate infusion for three days, terbutaUne 5 #g/h subcutaneously for nine days, and then oral terbutoline. The postoperative course was complicated by Clostridium diJ~cile pseudomembranous colitis from perioperative antibiotics but responded to a course of oral vancomycin. The mother and fetus had an
1 1 17
otherwise uneventful recovery; they left the hospital on the 14th day and returned to a small town in New Mexico to be followed by her local obstetrician. Postoperative ultrasound examinations showed normal amniotic fluid volume, and decompression of the urinary tract, but the fetus remained small for gestational age. At 32 weeks' gestation, during a snowstorm that prevented transport, the patient developed preterm labor that could not be controlled by tocolytic agents. A cesarean section delivery was uneventful, but the 1,280-g male had Apgar scores of 0 at one and five minutes, and was not resuscitated. The thorax was noted by the pediatrician in attendance to be very small even for the gestational age of 32 weeks. Autopsy showed deficient abdominal musculature, redundant abdominal skin, bilateral renal dysplasia, pulmonary hypoplasia, and urethral atresia. DISCUSSION
In cases of severe b i l a t e r a l hydronephrosis diagnosed in utero, the development of o l i g o h y d r a m n i o s before the third t r i m e s t e r g e n e r a l l y heralds a f a t a l n e o n a t a l outcome. 9 W h e n obstruction is severe enough to p r o d u c e o l i g o h y d r a m n i o s during the c a n a l i c u l a r stage of lung d e v e l o p m e n t (16 to 28 weeks' gestation), p u l m o n a r y h y p o p l a s i a a n d renal d y s p l a s i a ( T y p e IV cystic disease) p r e c l u d e n e o n a t a l survival. 1-1~'~6'18O u r e x p e r i m e n t a l work suggests t h a t the fatal p u l m o n a r y o u t c o m e m a y be reversed by early decompression of the o b s t r u c t e d u r i n a r y tract. 1,2,5,6 This initial small series of h u m a n cases confirms t h a t the d e v e l o p m e n t of f a t a l p u l m o n a r y hypoplasia can be prevented if a m n i o t i c fluid d y n a m i c s can be restored by decompression of the o b s t r u c t e d u r i n a r y t r a c t ( T a b l e 1). T h e t h r e e neonatal survivors h a d essentially n o r m a l p u l m o n a r y function at birth; all h a d reversal of o l i g o h y d r a m n i o s by intervention. It is likely t h a t all would have died from p u l m o n a r y h y p o p l a s i a without intervention. O f the two patients who died from hypoplastic lungs, one had persistent oligohyd r a m n i o s and the other had, in retrospect, clear evidence o f longstanding severe preexisting oligohydramnios. P a t i e n t s with this degree of preexisting oligohydramnios, ie, a d e q u a t e a m n i o t i c fluid never seen, should not in the future be considered for t r e a t ment. I t is not possible to ascertain from this s m a l l series w h e t h e r in-utero intervention a r r e s t e d or reversed cystic d y s p l a s t i e c h a n g e s c a u s e d by obstructive u r o p a t h y . It is possible t h a t the dysplastic changes initiated in utero will c o m p r o m i s e renal function progressively as functional d e m a n d increases with growth. Because h u m a n renal p a r e n c h y m a l d e v e l o p m e n t or maldevelo p m e n t is c o m p l e t e at birth, p o s t n a t a l relief o f obstruction m a y not prevent progression to end stage renal disease. 19"21 H e r e d i t a r y types of dysplasia a r e unlikely to respond to decompression in utero or postnatally, as t h e y result from d i s t u r b e d organogenesis prior to the eighth week o f gestation; and early, c o m p l e t e obstruc-
1 1 18
tion effects tubular and glomerular differentiation throughout the kidney resulting in irreversible dysplastic changes. However, the most common form of cystic dysplasia appears to be a developmental consequence of partial obstruction that affects cortical structures that develop later in gestation.2224Relief of obstruction during the most active phase of nephrogenesis, between 20 and 30 weeks' gestation, may obviate further damage and allow nephrogenesis to proceed normally. Experimentally, dysplastic changes similar to those seen in human neonates can be produced by obstruction early in gestation.3'25 Renal damage is related to the timing and duration of obstruction, and can be ameliorated by in-utero decompression. 3"4"26In sheep, reversal of urinary tract obstruction diminishes, but does not eliminate the renal damage observed at birth. 4 The long-term consequences of this residual renal damage have not been assessed either experimentally or clinically. Further experience and long-term followup are necessary to determine the effectiveness of in-utero decompression in reversing or arresting renal damage caused by obstruction. Recent diagnostic advances have greatly improved our ability to select from a large number of fetuses with dilated urinary tracts the rare fetus who requires intervention, ie, one at significant risk of pulmonary hypoplasia and renal dysplasia, but with sufficient residual renal function to restore amniotic fluid dynamics and support normal postnatal life. We have developed prognostic criteria (urine Na < 100 mEq/ L, CI < 90 mEq/L, osm < 210 mosm, and normal fetal kidneys by ultrasound) that reliably predict neonatal and long-term outcome.1~ In this early series, the two patients with a good outcome satisfied all the prognostic criteria (the elevated osmolarity in case 4 was felt to be a laboratory error), while the two who died and the one with renal failure failed to meet at least one criteria. Sonographic criteria are also useful. Increased echogenicity of renal parenchyma correlates with renal dysplasia,29 and both patients in this series with increased echogenicity had renal damage. Although echogenicity is subjective, it would be prudent to exclude from intervention a fetus with any poor prognostic factor including increased echogenicity. The etiology of the obstructive uropathy was posterior urethral value syndrome in two males and urethral atresia in the other. The two females had no anatomic basis for the obstruction and it was presumed to have a neurogenic origin. While it is possible that their functional obstruction may have spontaneously corrected in utero, the oligohydramnios was prolonged and severe, and neonatal survival would have been unlikely without restoration of amniotic fluid dynamics.
CROMBLEHOLME ET AL
The maternal risks of open fetal surgery remain a principle concern in electing this experimental treatment. Prior to offering fetal surgery to our patients, the efficacy of fetal decompression was established in sheep) "4 and the feasibility and safety of open fetal surgery were established in the more rigorous nonhuman primate mode.3~ The tocolytic, anesthetic, and surgical techniques were all developed and extensively tested in the nonhuman primate prior to this successful clinical application. Fortunately, in this early experience there were no intraoperative complications and the two postoperative complications, in retrospect, were both avoidable. The amniotic fluid leak in case 3 represented a technical error that prompted us to change the method of hysterotomy closure, and it has not recurred in subsequent cases. The mild case of pseudomenbranous colitis was secondary to excessively long perioperative antibiotic usage, and promptly responded to oral vancomycin. A feature common to each case was the problem of preterm labor that required continuous tocolytic therapy perioperatively and until delivery. Premature rupture of membranes occurred in two patients (cases 3 and 4). In case 3, delivery was successfully delayed by tocolytic agents for 2 weeks without infectious complications. In case 4, premature rupture of membranes prompted a cesarean section when labor could not be controlled with systemic tocolytic agents. The interval from fetal surgery to delivery ranged from 9 to 16 weeks, with a mean of 12 weeks. There were no long-term complications as a result of this procedure, and two patients have subsequently had successful normal pregnancies. We believe that the paucity of complications in these five patients and five others with open fetal procedures 33performed for other reasons, is a direct result of our extensive preparation in experimental fetal animals, which should be a prerequisite for anyone considering these inherently dangerous procedures. Because of improvements in selection criteria, prognosis for the fetus with bilateral hydronephrosis can now be realistically assessed and families counseled appropriately, s'1~ Most fetuses can be successfully managed after birth, but prenatal diagnosis of these cases allows improved postnatal care. The indications for intervention are narrow, and open fetal surgery is appropriate only for a select group of fetuses diagnosed in the late first and early second trimester with associated oligohydramnios, favorable fetal urine electrolytes and osmolarity, and normal renal parenchyma on ultrasound. Open fetal surgery is preferable to catheter decompression if a prolonged period (>2 to 3 weeks) remains until lung maturity, to allow reliable long-term decompression and restoration of amniotic fluid dynamics.
OPEN FETAL SURGERY FOR CONGENITAL HYDRONEPHROSIS
1 1 19
Open fetal surgery for congenital hydronephrosis is capable of decompressing the obstructed u r i n a r y tract, restoring amniotic fluid dynamics, a n d preventing otherwise fatal p u l m o n a r y hypoplasia. W h e t h e r inutero decompression reverses or arrests dysplastic changes in the kidneys and prevents long-term renal
failure r e m a i n s an open question. A l t h o u g h m a t e r n a l risk r e m a i n s a principle concern, hysterotomy for fetal surgery has thus far proven feasible a n d safe in experienced hands. Open decompression for fetal obstructive uropathy r e m a i n s a n experimental therapy awaiting controlled clinical trials to establish its efficacy.
REFERENCES
1. Harrison MR, Ross NA, Noall R, et al: Correction of congenital hydronephrosisI. The model: Fetal urethral obstruction produces hydronephrosis and pulmonary hypoplasia in fetal lambs. J Pediatr Surg 18:247-256, 1983 2. Harrison MR, Nakayama DK, NoaU R, et al: Correction of congenital hydronephrosis in utero II. Decompression reverses the effects of obstruction on the fetal lung and urinary tract. J Pediatr Snrg 17:965-974, 1982 3. Gliek PL, Harrison MR, Adziek NS, et al: Correction of congenital hydronephrosisin utero III. Early mid-trimester ureteral obstruction produces renal dysplasia. J Pediatr Surg 18:681-687, 1983 4. 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 5. Nakayama DK, Glick PL, Villa RL, et al: Experimental pulmonary hypoplasia, due to oligohydramniosand its reversal by relievingthoracic compression. J Pediatr Surg 18:347-353, 1983 6. Adzick NS, Harrison MR, Glick PL, et al: Pulmonary hypoplasia and oligohydramnios:Relative contributions of lung fluid and fetal breathing movements.J Pediatr Surg 19:658-665, 1984 7. Harrison MR, Filly RA, Parer JT, et al: Management of the fetus with a urinary tract malformation. JAMA 246:635-639, 1981 8. Golbus MS, Harrison MR, Filly RA, et al: In utero treatment of urinary tract obstruction. Am J Obstet Gynecol 142:383-386, 1982 9. Harrison MR, Golbus MS, Filly RA et al: Management of the fetus with congenital hydronephrosis. J Pediatr Surg 17:728-742, 1982 10. 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 11. Golbus MS, Harrison MR, Filly RA, et al: Prenatal diagnosis and treatment of fetal hydronephrosis.Semin Perinatol 7:102-105, 1983 12. Manning FA, Harrison MR, Rodeck CR, et al: Catheter shunts for fetal hydronephrosis and hydrocephalus:Report of the International Fetal Surgery Registry. N Engl J Med 315:336-340, 1986 13. Kramer SA: Current status of fetal intervention of congenital hydronephrosis. J Uroi 130:641-646, 1983 14. Berkowitz RL, Giickman MG, Smith GJW, et al: Fetal urinary tract obstruction: What is the role of surgical intervention in utero? Am J Obstet Gynecol 144:367-375, 1982 15. McFadyen IR, Wiggleworth JS, Dillon MJ: Fetal urinary tract obstruction: Is active interventionbefore deliveryindicated? Br J Obstet Gynecoi 90:342-349, 1983 16. Harrison MR, Goibus MS, Filly RA, et al: Fetal surgery for congenital hydronephrosis.N Engi J Med 306:591-593, 1982 17. Harrison MR, Golbus MS, Filly RA, et al: Fetal hydroneph-
rosis: Selection and surgical repair. J Pediatr Surg 22:556-558, 1987 18. Nakayama DK, Harrison MR, de Lorimier AA: Prognosisof posterior urethral valves presenting at birth. J Pediatr Surg 21:4345, 1986 19. Warshaw BL, Edelbrock HH, Ettinger RB: Progression to end stage renal disease in children with obstructive uropathy. J Pediatr 100:182-187, 1982 20. MeCrory WW: Developmental Nephrology. Harvard University Press, Cambridge, MA, 1972, pp 40-46 21. OliverJ: Nephrons and Kidneys. Harper & Row, New York, 1972, pp 1-24 22. Bernstein J: A classification of renal cysts, in Gardner KD (ed): Cystic Diseaseof the Kidney.New York, John Wiley and Sons, 1976, pp 7-30 23. Bernstein J: The morphogenesisof renal parenchymal maldevelopment (renal dysplasia). Pediatr Clin North Am 18:395-407, 1971 24. Potter EL: Normal and Abnormal Developmentof the Kidney. Chicago, Year Book Medical, 1972, pp 209-221 25. Steinhardt GF, Vogler G, Salinas-Madrigal L, et al: Induced renal dysplasiain the young pouchopossum. J Pediatr Surg 23:11271130, 1988 26. Beck AD: The effect of intrauterine urinary Obstructionupon developmentof the fetalkidney. J Urol 105:784-789, 1981 27. CrombleholmeTM, Harrison MR, Longaker MT, et al: Fetal interventionfor bilateral hydronephrosis:Patient selectionand therapeutic etticaey. (submitted) 28. Inselman LS, MeUins RB: Growth and development of the lung. J Pediatr 98:1-15, 1981 29. Mahoney BS, Filly RA, Callen PW, et al: Sonographic evaluation of fetal renal dysplasia. Radiology 152:143-146, 1984 30. Harrison MR, Anderson J, Rosen MS: Fetal surgery in the primate I. Anesthetic, surgical and toeolytie management to maximize fetal-neonatal survival.J Pediatr Surg 17:115-122, 1982 31. Nakayama DK, Harrison MR, Seron-Ferre M, et al: Fetal surgery in the primate II. Uterine electromyographie response to operative procedures and pharmacologic agents. J Pediatr Surg 19:333-339, 1984 32. Adzick NS, Harrison MR, Glick PL, et al: Fetal surgery in the primate III. Maternal outcome after fetal surgery. J Pediatr Snrg 21:477-480, 1986 33. Harrison, MR: Unpublished data. 34. Harrison MR, Golbus MS, Filly RA (eds): The Unborn Patient. Philadelphia, Grune & Stratton, 1984, pp 277-348 35. Glazer GM, Filly RA, Callen PW: The varied sonographic appearance of the urinary tract in the fetus and newborn with urethral obstruction. Radiology 144:563-568, 1982 36. Hobbins JC, Romero R, Gronnum P, et al: Antenatal diagnosis of renal anomalies with ultrasound I. Obstructive uropathy. Am J Obstet Gynecol 148:868-877, 1984
1120
CROMBLEHOLME ET AL
Discussion Stephen Dunn (Philadelphia): The lethal defect in these children is pulmonary hypoplasia. The fact that some of these children may not develop normal renal function doesn't seem to me to be such a bad thing. You are now salvaging children who have normal intelligence, with normal developmental capabilities otherwise, and allowing them to get to the point where they may undergo a renal transplant. I think all of the information that we have now would say that the prognosis with renal transplantation for a child developing renal failure in this age group is very good. I congratulate you on your results and I think that you should not stop performing this procedure even if you cannot achieve completely normal renal function. Correction of pulmonary hypoplasia is an important accomplishment in these children. M. Dimler (Phoenix): I would like to ask one question concerning the degree of dysplasia and its relationship to timing of intervention particularly with your non-human primate model. Do you have data that are helpful in sorting out the time course at which such severe dysplasia is present that one should not contemplate salvage? T. Weber (St Louis): Extracorporeal membrane oxygenation has redefined lethal pulmonary hypoplasia. In our series of 130 newborns treated with ECMO, we have two babies born with severe oligohydramnios, bell-shaped chest, and pulmonary hypoplasia; they were treated with ECMO and both have survived. I think at least some of the pathophysiology in these newborns is pulmonary hypertension rather than insufficient pulmonary parenchyma. We need to look at aggressive means of support for these babies, and they should still be considered candidates for ECMO. J. Noseworthy (Oncinnati): Certainly the San Francisco group continues to push us and challenge us to think about new methods of dealing with diseased lungs and kidneys. Would you comment about patientSelection despite the fact that you want that dealt with a little bit later on? I think it is an important aspect of what you have presented. Could you give us some general outlines? J. Langer (response): Dr Dunn, we would agree wholeheartedly with your comments. The field of transplantation is still in its infancy, and we are sure that further advances will allow children with chronic renal failure to have better qualities of life. Dr Dimler, it is always difficult to apply an animal model to the human situation, but certainly there seems to be a very strong relationship between the amount of time that the urinary tract is obstructed and the degree of renal damage. We do not know how many
weeks' gestation that translates to in the human fetus, and this underlines the importance of developing prognostic criteria that we can trust. One of these may be the gestational age at which hydronephrosis or oligohydramnios occurs. Dr Noseworthy, I think that there are two issues to deal with here. One is pulmonary outcome and the other is renal outcome, and when we look at prognostic criteria and selection, we have to think of them separately. It is now clear to us that the presence of early oligohydramnios is a very bad prognostic sign for pulmonary hypoplasia. The fetus who has oligohydramnios without evidence of longstanding pre-existing oligohydramnios or in particular the fetus who has hydronephrosis and then develops oligohydramnios, is the kind of patient at risk for pulmonary hypoplasia and should have in utero intervention. The prognostic criteria for the ultimate development of chronic renal failure is a far more difficult question. The things that we have looked at are fetal urinary electrolytes and osmolarity, ultrasound evidence of dysplasia in the kidneys, as well as the presence of oligohydramnios. We are well on our way to determining how to interpret these data, but we still have a long way to go. T. Bartholomew (San Antonio, TX): You have not mentioned anything about the complications of your intervention. Those are real; you pointed out one abortion from chorioamnitis, and also you haven't pointed out the accuracy of your diagnoses. These patients may have prune belly syndrome; in some of your drawings, in fact used on your slides, a patient is depicted with prune belly syndrome with an assumption that the intravesicle obstruction in some way caused the prue belly syndrome. I don't think that can be accepted. Do you have any comment on the accuracy of the diagnoses that have been involved and the complications of your intervention? T.M. Crombleholme (response): Dr Bartholomew, in terms of the complications, three vesicoamniotic catheter shunts were complicated by chorioamnioitis. These occurred early in our experience, prior to the use of prophylactic antibiotics. These have been no maternal complications as a result of their placement. In open fetal surgery, there have been two complications: pseudomembranous colitis and a single instance of amniotic fluid leak due to a technical problem that has been rectified and has not recurred. Stephen Dunn (Philadelphia): How do you choose either the Pigtail Harrison catheter or an open fetal procedure? How do your criteria assist you in making those decisions? The second comment I had was that I think your data on urine electrolytes is interesting, but
OPEN FETAL SURGERY FOR CONGENITAL HYDRONEPHROSIS
in using the electrolyte data as one of your criteria for separating good from poor prognosis groups, you expect them to have a significant difference in the final analysis in the first place. T.M. Crombleholme (response): Dr Dunn, you raise a crucial issue and one with which we continually struggle. We have not yet refined the criteria sufficiently to distinguish those patients in the poor prognostic group who will survive but go on to have renal failure from those with no hope of survival. S. Kramer (Rochester, NY): You have shown nicely the difference between the good and poor prognostic groups. You have not convinced me yet, though, that the good prognostic group would not do just as well if you left them alone and then did a vesicostomy or ureterostomy at birth. In the six patients you showed that did well with intervention, don't you think perhaps they would have done just as well if they were delivered at term and then had vesicostomies at that time? We have had that experience and there is a report in the Journal of Urology (May issue) of a patient with
1121
oligohydramnios and bilateral hydronephrosis who was delivered, had a vesicostomy, and now at 2 years of age, has normal pulmonary and normal renal function. T.M. Crombleholme (response): Dr Kramer, with regard to the issue of whether or not the good prognosis patients would have done well without intervention, six of the eight patients who survived had oligohydramnios, and would not have survived without restoration of amniotic fluid. Survival of patients with oligohydramnios diagnosed early remains anecdotal. T.M. Crombleholme (closing): The decision to use a vesicoamniotic shunt as opposed to an open procedure depends largely upon the time at which the patient presents. The functional life span of a vesicoamniotic shunt is less than 2 weeks. A patient who presents early in gestation (at 20 to 24 weeks) would require multiple catheter placements which are not without complications. In that instance, we would recommend open fetal surgery.
9 The fetus w i t h severe bilateral hydronephrosis and 9 a s s o c i a t e d oligohydramnios in t h e s e c o n d t r i m e s t e r i s . d o o m e d a t b i r t h by ongoing p u l m o n a r y a n d renal d a m a g e . Since decompression w i t h percutaneously placed catheters is not dependable for long periods, w e have developed anesthetic, surgical, and tocolytic techniques for open fetal 9 urinary tract decompression in animals, and have n o w applied those techniques to a small group of five patients.. One had bilateral ureterostomies and the subsequent f o u r , had marsupialization of the bladder. All pregnancies proceeded to cesarean delivery at 32 to 35 weeks" gestation. There was no long-term maternal morbidity, and t w o mothers have since experienced normal pregnancies. Three fetuses had return of normal amniotic fluid dynamics, and all three had adequate pulmonary function at birth, suggesting~that fatal pulmonary hypoplasia associated w i t h early severe oligohydramnios had boon reversed. ~ w o neonates died at birth w i t h pulmonary~ hypoplasia. One had no amniotic fluid even after decompression, and the other had some amniotic fluid after decompression but a tiny chest cavity due to the long period of severe oligohydramnios before decompression. Of the three surviving infants, one had normal renal func-. tion when she died of unrelated causes at 9 months of age. One has normal renal function at 23 months and the third had failing renal function at 21/2 years and has grown and developed normally, but will require renal transplantation. We have n o w developed selection criteria that would exclude from t r e a t m e n t the t w o fetuses w h o died of pulmonary hypoplasia and the one w h o developed renal 9 failure. Our initial limited experience suggests the following: (1) that open fetal surgery is safe in experienced hands and the risks of open fetal surgery can be minimized by extensive prepatory experimental w o r k ; (2) that in a f e w highly selected cases, open decompression can restore amniotic fluid dynamics and is efficacious in preventing pulmonary hypoplasia at birth; and (3)' that the effect of decompression on the development of renal dysplasia and. ult.mate renal function remahs unknown. 9 1988 b y (;rune & Stratton, inc.
been established by many experimental studies? "6 These studies demonstrate that potentially fatal pulmonary hypoplasia, caused by oligohydramnios resulting from obstructive uropathy, may be ameliorated by decompression in utero. 1'2'5'6 In addition, in-utero decompression may arrest dysplastic morphogenesis of the kidney and preserve renal function depending on the timing and the duration of the obstruction) "4 The efficacy of this procedure in human fetuses is yet to be established9 Vesicoamniotic shunts have been employed in the short-term treatment of bilateral fetal hydronephrosis, but due to the high incidence of catheter clogging or displacement and the associated risk of chorioamnionitis, vesicoamniotic shunts are not satisfactory for longterm fetal urinary tract decompression. 7"~2 Severely affected fetuses with associated oligohydramnios diagnosed prior to 24 weeks' gestation generally have not survived even when successfully decompressed. 9"~7Due to improvements in prognostic criteria, we may now be able to select for intervention those fetuses whose kidneys are not already irreversibly damaged and could, if decompressed, produce sufficient urine to restore amniotic fluid dynamics, preventing neonatal death from pulmonary hypoplasia. This brief review is a report of our initial experience with open fetal surgery performed in five cases of bilateral hydronephrosis. These cases illustrate that open fetal surgery is feasible with acceptable maternal risk, and that restoration of normal amniotic fluid dynamics can reverse potentially fatal pulmonary hypoplasia. Efficacy in reversing renal dysplasia is not known. MATERIALS AND METHODS
INDEX WORDS: Fetal surgery; congenital hydronephrosis.
T
HE THEORETICAL BASIS for in-utero decompression of bilateral fetal hydronephrosis has
From the Fetal Treatment Program and the Departments of Surgery, Obstetrics, Gynecology and Reproductive Science, and Radiology, University of California, San Francisco. Presented at the 19th Annual Meeting of the American Pediatric Surgical Association, Tucson, Arizona, May I 1-14, 1988. Address reprint requests to Michael R. Harrison. MD. The Fetal Treatment Program, 585 HSE, UCSF, Third and Parnassas Ave, San Francisco, CA 94143. 9 1988 by Grune & Stratton, Inc. 0022-3468/88/2312-0005503.00/0 1114
More than 200 cases of bilateral fetal hydronephrosis have been referred to the Fetal Treatment Program at the University of California, San Francisco Medical Center from 1978 to 1988. Five patients were found suitable for open intervention. Two of these cases have been previously reported. ]6']7All surgical procedures were performed by M.R.H.
CASE REPORTS
Case 1 A primagravida woman was referred with a male fetus (G.T.) of 20 weeks' gestation who had oligohydramnios and bilateral congenital hydronephrosis. Fetal bilateral ureterostomies were created, but oligohydranmios persisted and the neonate died of pulmonary hypoplasia. '~
Journal of Pediatric Surgery, Vo123, No 12 (December),1988: pp 1114-1121
OPEN FETAL SURGERY FOR CONGENITAL HYDRONEPHROSIS
1115
Case 2 A 31-year-old primagravida woman was referred from Texas with a male fetus of 22 weeks' gestation. The fetus (M.M.) had oligohydramnios and bilateral hydronephrosis. Open fetal surgery was successfully performed at 24 weeks' gestation with decompression of the urinary tract and restoration of amniotic fluid volume for the remainder of the gestation. The infant had no significant respiratory difficulty and has developed normally; however, renal function has deteriorated gradually. Although dialysis has not been necessary (the boy is now 2V2years old), transplantation is planned) 7
B
Pulse imeter
Case 3 A 19-year-old primagravida woman was referred from Minnesota at 22 weeks' gestation. The female fetus (M.S.) bad gross megacystis with severe bilateral hydronephrosis and slightly ecbogenic renal parenchyma. The patient's amniotic fluid volume at presentation was low and then disappeared entirely. Percutaneous fetal bladder drainage for four hours showed good urine production (4 mL/h) with a sodium level of 79 mEq/L, a chloride level of 71 mEq/L, and osmolarity of 150 mosm. This electrolyte pattern suggested good potenttal for functional renal recovery, l~ Karyotype was normal, 46XX. Over five days, the oligohydramnios continued to worsen and fetal bladder distention increased. The family was counseled about the options available including termination, observation, vesicoamniotic catheter shunt, and open fetal surgery. After extensive discussion of the risks of hysterotomy and fetal bladder marsupialization, and the experimental nature of the procedure, the family chose to undergo open decompression. The surgery was performed at 23 weeks' gestation using anesthetic and tocolytic regimens previuosly described. 16.17Briefly, this included indomethacin, haiothane for uterine relaxation, fetal and maternal anesthesia intraoperatively, and ritodrine infusion and/or indomethacin postoperatively. The uterus was exposed by a low transverse abdominal incision, and intraoperative sonography confirmed the fetal position as breech facing anterior and the placenta was posterior. A low transverse hysterotomy was made directly over the massively distended fetal abdomen and bladder, and due to the complete absence of amniotic fluid, the fetal abdomen was entered immediately. The bysterotomy was extended with a newly developed resorptive stapling device and the lower half of the fetus exterionized. The massively distended bladder was opened and sutured to the thinned out redundant abdominal wall with 3-0 silk, thus scaling the peritoneal cavity and marsupializing the bladder (Fig 1). The fetus was returned, with some difficulty due to the small amniotic cavity, and the uterus was dosed with all layer interrupted Dexan over a running Dexon inner layer. Amniotic fluid volume was restored with warm Ringer's lactate containing oxacillin. Fetal heart rate monitored sonographically remained remarkably stable throughout the procedure. The postoperative course was uneventful. Ritodrine was given intravenously for three days and then orally. The patient was discharged after nine days and returned to Minnesota. The fetal condition improved dramatically as serial ultrasound documented complete decompression of the fetal megacystis, hydronephrosis with minimal residual caliectasis, and normal amniotie fluid volume. The thin redundant abdominal wall ("prune belly") improved and interval growth was normal. Twenty-four days postoperatively the mother presented with abdominal pain peritoneal signs, and preterm labor. At laparotomy, her obstetrician found a small amniotic fluid leak from the hysterotomy incision which was easily controlled with a single O chromic suture. She was again discharged on oral tocolytics. At 31 weeks' gestation, she was readmitted with spontaneous premature rupture of membranes and maintained on oral ritodrine with bed rest, without evidence of
C
Fig 1. Example of open fetal surgery for hydronephrosis. IA) Bilateral hydronephrosis, megscystis, and oligohydramnios at approximately 24 weeks" gestation. Guided by intrsoperative sonography, a hysterotomy is placed to avoid the placenta and expose the fetus" lower body without disturbing the umbilical circulation. (B) Fetal heart rate and oxygen saturation are monitored by transcutaneous pulse oximeter and ECG while a cutaneous vesicostomy is constructed. (C) Amniotic fluid volume restored by fetal urine. 17
chorioanmionitis for 2 weeks. A cesarean section was performed at 33 weeks' gestation; the previous hysterotomy incision was found to be thinned but intact. The 4 lb 2 oz baby girl required minimal ventilatory support. Although her respiratory function was normal after birth, she required neonatal exploration and ileal diversion for mieroileum and microcolon. Her renal function remained normal, but persistent gastrointestinal malfunction suggested pseudoobstruction, which did not improve despite further surgical intervention and eventually led to death from septicemia at 9 months of age.
Case 4 This 35-year-old prirnagravida woman was found at 16 weeks' gestation to have oligohydramnios, and the female fetus (CH) had megacystis and bilateral hydronepbrosis. Ultrasound showed no evidence of cystic renal parenchymal changes. Aspirated fetal urine sodium concentration of 100 mEq/L and chloride concentration of 91 mEq/L suggested good potential for renal recovery, although osmolarity was slightly elevated at 221 mosm. Percutaneous umbilical blood sampling (PUBS) for karyotype analysis yielded a normal 46XX result. The family who had battled infertility for years, was counseled at length about all the available options. They resolutely refused termination and insisted on pursuing treatment, although we were reluctant to offer treatment to a female fetus since the pathophysioiogy was not clearly defined. At 18 weeks' gestation, open fetal surgery was performed as described in the previous cases. Intraoperative sonography showed the fetus in a transverse position wffh the lower extremities to the : left. Unfortunately, a placentae previa covered the entire anterior
1 1 16
CROMBLEHOLME ET AL
o 8
.o Z
z=
"~2
;I
o
r~
E~
E~
E~ c
iZ
| -~
0
!
~
o o
,~
o o
~
-
o
o
~
~
E
~-~
~ E LUre E E~
~ o'~ "e~ o ~ E ~ ~ E ~ 0 E~ Emm = e9 E~O E~
z~o
z~oz~oz~o
~.~
~
|
o
co
0
"o
,==.~.-=
~
o
~"
"o
|
|
>-
|
~
0
o
0
=~._~
|
~'6
>'-
=-
|
>
=
-
~ ~>..;
6 R "o
.>
OPEN FETAL SURGERY FOR CONGENITAL HYDRONEPHROSIS
wall of the uterus from cervix to fundus. The uterus was opened at the only placenta-free area on the left posterior wall of the fundus. The lower extremities of the fetus were exteriorized, and a transcutaneous pulse oximeter was placed around the fetal thigh. A midline suprapubie incision was made through the fetal abdomen, exposing a distended thick walled bladder.The bladder was opened and rnarsupializedto the fetalabdominal wall with interrupted4-0 silksutures. The fetus was returned to the uterine cavity within ten minutes. Amniotic fluid was restored with warm normal saline.The uterus was closed with interruptedO - P D S all layer sutures. After uterine closure, sonography showed a strong fetal heart rate of 140. Ritodrine hydrochloride infusionwas weaned within three days, and the patient was switched to oral terbutoline for control of uterine irritability.Both the mother and fetus recovered uneventfully.The patientleftthe hospitalaftereightdays, and returned to her home in southern California. The fetus grew normally. The bladder remained decompressed, and was normal upon subsequent ultrasound examinations. Ten
weeks after fetal surgery, at 28 weeks' gestation, the patient presented to her local medical center with vaginal bleeding and preterm labor thought to be due to mild abruptio placentae. The bleeding stopped with conservative measures, and preterm labor was controlled for the next 19 days with bed rest, ritodrine, and occasionally magnesium sulfate. The membranes ruptured at 32 weeks' gestation and a cesarean section was performed. The uterine incision was well healed. A healthy 1,750-g baby girl had Apgar scores of 9 at one minute and 9 at five minutes. Ultrasound showed a distended bladder and hydronephrosis but normal renal parenehyma. Neonatal creatinine was 1.1 mg/dL, but at 1 month fell to and remained at 0.6 mg/dL. Voiding cystourethrogram showed a hypertrophied bladder with dynamic evidence of neurogenic dysfunction and an elongated nonobstructed urethra. The infant's course has been uncomplicated except for a urinary tract infection at 3 weeks of age. She has continued to do well with normal renal function, and is now 2 years old.
Case 5 A 27-year-old gravida 4 para 3 mother was referred from New Mexico at 20 weeks' gestation with oligohydramnios, and a small (for gestational age) fetus (K.G.) with massive fetal abdominal distention, megaeystis, and bilateral hydronephrosis. Karyotype analysis yielded a normal (46XY) result. Both kidneys evidenced increased eehogenicity but no cortical cysts, and aspirated fetal urine had borderline values (Na, 100 mEq/L; CI, 82 mEq/L; osm, 215 mosm/kg). The family was counseled about the options available and the likelihood of a poor outcome regardless of the course chosen. The family adamantly refused termination and requested open fetal surgery despite the risks of an experimental procedure and the guarded prospects for a successful outcome. At 22 weeks' gestation, open fetal surgery was performed using the regimen described above. Intraoperative sonography showed the fetus was lying head down, with back anterior. Since the placenta was anterior near the top of the fundus, the hysterotomy was made posteriorly. The lower extremities were delivered into the wound and a transcutaneous pulse oximeter was applied to the fetal thigh. The vesicostomy was performed as in cases 2 to 4, and required 12 minutes of extrauterine exposure. The hysterotomy was closed with through-and-through O-PDS mattress sutures, and two layers of running Dexon. The amniotic fluid volume was restored with warm normal saline. Postoperative tocolysis included indomethaein per rectum, magnesium sulfate infusion for three days, terbutaUne 5 #g/h subcutaneously for nine days, and then oral terbutoline. The postoperative course was complicated by Clostridium diJ~cile pseudomembranous colitis from perioperative antibiotics but responded to a course of oral vancomycin. The mother and fetus had an
1 1 17
otherwise uneventful recovery; they left the hospital on the 14th day and returned to a small town in New Mexico to be followed by her local obstetrician. Postoperative ultrasound examinations showed normal amniotic fluid volume, and decompression of the urinary tract, but the fetus remained small for gestational age. At 32 weeks' gestation, during a snowstorm that prevented transport, the patient developed preterm labor that could not be controlled by tocolytic agents. A cesarean section delivery was uneventful, but the 1,280-g male had Apgar scores of 0 at one and five minutes, and was not resuscitated. The thorax was noted by the pediatrician in attendance to be very small even for the gestational age of 32 weeks. Autopsy showed deficient abdominal musculature, redundant abdominal skin, bilateral renal dysplasia, pulmonary hypoplasia, and urethral atresia. DISCUSSION
In cases of severe b i l a t e r a l hydronephrosis diagnosed in utero, the development of o l i g o h y d r a m n i o s before the third t r i m e s t e r g e n e r a l l y heralds a f a t a l n e o n a t a l outcome. 9 W h e n obstruction is severe enough to p r o d u c e o l i g o h y d r a m n i o s during the c a n a l i c u l a r stage of lung d e v e l o p m e n t (16 to 28 weeks' gestation), p u l m o n a r y h y p o p l a s i a a n d renal d y s p l a s i a ( T y p e IV cystic disease) p r e c l u d e n e o n a t a l survival. 1-1~'~6'18O u r e x p e r i m e n t a l work suggests t h a t the fatal p u l m o n a r y o u t c o m e m a y be reversed by early decompression of the o b s t r u c t e d u r i n a r y tract. 1,2,5,6 This initial small series of h u m a n cases confirms t h a t the d e v e l o p m e n t of f a t a l p u l m o n a r y hypoplasia can be prevented if a m n i o t i c fluid d y n a m i c s can be restored by decompression of the o b s t r u c t e d u r i n a r y t r a c t ( T a b l e 1). T h e t h r e e neonatal survivors h a d essentially n o r m a l p u l m o n a r y function at birth; all h a d reversal of o l i g o h y d r a m n i o s by intervention. It is likely t h a t all would have died from p u l m o n a r y h y p o p l a s i a without intervention. O f the two patients who died from hypoplastic lungs, one had persistent oligohyd r a m n i o s and the other had, in retrospect, clear evidence o f longstanding severe preexisting oligohydramnios. P a t i e n t s with this degree of preexisting oligohydramnios, ie, a d e q u a t e a m n i o t i c fluid never seen, should not in the future be considered for t r e a t ment. I t is not possible to ascertain from this s m a l l series w h e t h e r in-utero intervention a r r e s t e d or reversed cystic d y s p l a s t i e c h a n g e s c a u s e d by obstructive u r o p a t h y . It is possible t h a t the dysplastic changes initiated in utero will c o m p r o m i s e renal function progressively as functional d e m a n d increases with growth. Because h u m a n renal p a r e n c h y m a l d e v e l o p m e n t or maldevelo p m e n t is c o m p l e t e at birth, p o s t n a t a l relief o f obstruction m a y not prevent progression to end stage renal disease. 19"21 H e r e d i t a r y types of dysplasia a r e unlikely to respond to decompression in utero or postnatally, as t h e y result from d i s t u r b e d organogenesis prior to the eighth week o f gestation; and early, c o m p l e t e obstruc-
1 1 18
tion effects tubular and glomerular differentiation throughout the kidney resulting in irreversible dysplastic changes. However, the most common form of cystic dysplasia appears to be a developmental consequence of partial obstruction that affects cortical structures that develop later in gestation.2224Relief of obstruction during the most active phase of nephrogenesis, between 20 and 30 weeks' gestation, may obviate further damage and allow nephrogenesis to proceed normally. Experimentally, dysplastic changes similar to those seen in human neonates can be produced by obstruction early in gestation.3'25 Renal damage is related to the timing and duration of obstruction, and can be ameliorated by in-utero decompression. 3"4"26In sheep, reversal of urinary tract obstruction diminishes, but does not eliminate the renal damage observed at birth. 4 The long-term consequences of this residual renal damage have not been assessed either experimentally or clinically. Further experience and long-term followup are necessary to determine the effectiveness of in-utero decompression in reversing or arresting renal damage caused by obstruction. Recent diagnostic advances have greatly improved our ability to select from a large number of fetuses with dilated urinary tracts the rare fetus who requires intervention, ie, one at significant risk of pulmonary hypoplasia and renal dysplasia, but with sufficient residual renal function to restore amniotic fluid dynamics and support normal postnatal life. We have developed prognostic criteria (urine Na < 100 mEq/ L, CI < 90 mEq/L, osm < 210 mosm, and normal fetal kidneys by ultrasound) that reliably predict neonatal and long-term outcome.1~ In this early series, the two patients with a good outcome satisfied all the prognostic criteria (the elevated osmolarity in case 4 was felt to be a laboratory error), while the two who died and the one with renal failure failed to meet at least one criteria. Sonographic criteria are also useful. Increased echogenicity of renal parenchyma correlates with renal dysplasia,29 and both patients in this series with increased echogenicity had renal damage. Although echogenicity is subjective, it would be prudent to exclude from intervention a fetus with any poor prognostic factor including increased echogenicity. The etiology of the obstructive uropathy was posterior urethral value syndrome in two males and urethral atresia in the other. The two females had no anatomic basis for the obstruction and it was presumed to have a neurogenic origin. While it is possible that their functional obstruction may have spontaneously corrected in utero, the oligohydramnios was prolonged and severe, and neonatal survival would have been unlikely without restoration of amniotic fluid dynamics.
CROMBLEHOLME ET AL
The maternal risks of open fetal surgery remain a principle concern in electing this experimental treatment. Prior to offering fetal surgery to our patients, the efficacy of fetal decompression was established in sheep) "4 and the feasibility and safety of open fetal surgery were established in the more rigorous nonhuman primate mode.3~ The tocolytic, anesthetic, and surgical techniques were all developed and extensively tested in the nonhuman primate prior to this successful clinical application. Fortunately, in this early experience there were no intraoperative complications and the two postoperative complications, in retrospect, were both avoidable. The amniotic fluid leak in case 3 represented a technical error that prompted us to change the method of hysterotomy closure, and it has not recurred in subsequent cases. The mild case of pseudomenbranous colitis was secondary to excessively long perioperative antibiotic usage, and promptly responded to oral vancomycin. A feature common to each case was the problem of preterm labor that required continuous tocolytic therapy perioperatively and until delivery. Premature rupture of membranes occurred in two patients (cases 3 and 4). In case 3, delivery was successfully delayed by tocolytic agents for 2 weeks without infectious complications. In case 4, premature rupture of membranes prompted a cesarean section when labor could not be controlled with systemic tocolytic agents. The interval from fetal surgery to delivery ranged from 9 to 16 weeks, with a mean of 12 weeks. There were no long-term complications as a result of this procedure, and two patients have subsequently had successful normal pregnancies. We believe that the paucity of complications in these five patients and five others with open fetal procedures 33performed for other reasons, is a direct result of our extensive preparation in experimental fetal animals, which should be a prerequisite for anyone considering these inherently dangerous procedures. Because of improvements in selection criteria, prognosis for the fetus with bilateral hydronephrosis can now be realistically assessed and families counseled appropriately, s'1~ Most fetuses can be successfully managed after birth, but prenatal diagnosis of these cases allows improved postnatal care. The indications for intervention are narrow, and open fetal surgery is appropriate only for a select group of fetuses diagnosed in the late first and early second trimester with associated oligohydramnios, favorable fetal urine electrolytes and osmolarity, and normal renal parenchyma on ultrasound. Open fetal surgery is preferable to catheter decompression if a prolonged period (>2 to 3 weeks) remains until lung maturity, to allow reliable long-term decompression and restoration of amniotic fluid dynamics.
OPEN FETAL SURGERY FOR CONGENITAL HYDRONEPHROSIS
1 1 19
Open fetal surgery for congenital hydronephrosis is capable of decompressing the obstructed u r i n a r y tract, restoring amniotic fluid dynamics, a n d preventing otherwise fatal p u l m o n a r y hypoplasia. W h e t h e r inutero decompression reverses or arrests dysplastic changes in the kidneys and prevents long-term renal
failure r e m a i n s an open question. A l t h o u g h m a t e r n a l risk r e m a i n s a principle concern, hysterotomy for fetal surgery has thus far proven feasible a n d safe in experienced hands. Open decompression for fetal obstructive uropathy r e m a i n s a n experimental therapy awaiting controlled clinical trials to establish its efficacy.
REFERENCES
1. Harrison MR, Ross NA, Noall R, et al: Correction of congenital hydronephrosisI. The model: Fetal urethral obstruction produces hydronephrosis and pulmonary hypoplasia in fetal lambs. J Pediatr Surg 18:247-256, 1983 2. Harrison MR, Nakayama DK, NoaU R, et al: Correction of congenital hydronephrosis in utero II. Decompression reverses the effects of obstruction on the fetal lung and urinary tract. J Pediatr Snrg 17:965-974, 1982 3. Gliek PL, Harrison MR, Adziek NS, et al: Correction of congenital hydronephrosisin utero III. Early mid-trimester ureteral obstruction produces renal dysplasia. J Pediatr Surg 18:681-687, 1983 4. 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 5. Nakayama DK, Glick PL, Villa RL, et al: Experimental pulmonary hypoplasia, due to oligohydramniosand its reversal by relievingthoracic compression. J Pediatr Surg 18:347-353, 1983 6. Adzick NS, Harrison MR, Glick PL, et al: Pulmonary hypoplasia and oligohydramnios:Relative contributions of lung fluid and fetal breathing movements.J Pediatr Surg 19:658-665, 1984 7. Harrison MR, Filly RA, Parer JT, et al: Management of the fetus with a urinary tract malformation. JAMA 246:635-639, 1981 8. Golbus MS, Harrison MR, Filly RA, et al: In utero treatment of urinary tract obstruction. Am J Obstet Gynecol 142:383-386, 1982 9. Harrison MR, Golbus MS, Filly RA et al: Management of the fetus with congenital hydronephrosis. J Pediatr Surg 17:728-742, 1982 10. 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 11. Golbus MS, Harrison MR, Filly RA, et al: Prenatal diagnosis and treatment of fetal hydronephrosis.Semin Perinatol 7:102-105, 1983 12. Manning FA, Harrison MR, Rodeck CR, et al: Catheter shunts for fetal hydronephrosis and hydrocephalus:Report of the International Fetal Surgery Registry. N Engl J Med 315:336-340, 1986 13. Kramer SA: Current status of fetal intervention of congenital hydronephrosis. J Uroi 130:641-646, 1983 14. Berkowitz RL, Giickman MG, Smith GJW, et al: Fetal urinary tract obstruction: What is the role of surgical intervention in utero? Am J Obstet Gynecol 144:367-375, 1982 15. McFadyen IR, Wiggleworth JS, Dillon MJ: Fetal urinary tract obstruction: Is active interventionbefore deliveryindicated? Br J Obstet Gynecoi 90:342-349, 1983 16. Harrison MR, Goibus MS, Filly RA, et al: Fetal surgery for congenital hydronephrosis.N Engi J Med 306:591-593, 1982 17. Harrison MR, Golbus MS, Filly RA, et al: Fetal hydroneph-
rosis: Selection and surgical repair. J Pediatr Surg 22:556-558, 1987 18. Nakayama DK, Harrison MR, de Lorimier AA: Prognosisof posterior urethral valves presenting at birth. J Pediatr Surg 21:4345, 1986 19. Warshaw BL, Edelbrock HH, Ettinger RB: Progression to end stage renal disease in children with obstructive uropathy. J Pediatr 100:182-187, 1982 20. MeCrory WW: Developmental Nephrology. Harvard University Press, Cambridge, MA, 1972, pp 40-46 21. OliverJ: Nephrons and Kidneys. Harper & Row, New York, 1972, pp 1-24 22. Bernstein J: A classification of renal cysts, in Gardner KD (ed): Cystic Diseaseof the Kidney.New York, John Wiley and Sons, 1976, pp 7-30 23. Bernstein J: The morphogenesisof renal parenchymal maldevelopment (renal dysplasia). Pediatr Clin North Am 18:395-407, 1971 24. Potter EL: Normal and Abnormal Developmentof the Kidney. Chicago, Year Book Medical, 1972, pp 209-221 25. Steinhardt GF, Vogler G, Salinas-Madrigal L, et al: Induced renal dysplasiain the young pouchopossum. J Pediatr Surg 23:11271130, 1988 26. Beck AD: The effect of intrauterine urinary Obstructionupon developmentof the fetalkidney. J Urol 105:784-789, 1981 27. CrombleholmeTM, Harrison MR, Longaker MT, et al: Fetal interventionfor bilateral hydronephrosis:Patient selectionand therapeutic etticaey. (submitted) 28. Inselman LS, MeUins RB: Growth and development of the lung. J Pediatr 98:1-15, 1981 29. Mahoney BS, Filly RA, Callen PW, et al: Sonographic evaluation of fetal renal dysplasia. Radiology 152:143-146, 1984 30. Harrison MR, Anderson J, Rosen MS: Fetal surgery in the primate I. Anesthetic, surgical and toeolytie management to maximize fetal-neonatal survival.J Pediatr Surg 17:115-122, 1982 31. Nakayama DK, Harrison MR, Seron-Ferre M, et al: Fetal surgery in the primate II. Uterine electromyographie response to operative procedures and pharmacologic agents. J Pediatr Surg 19:333-339, 1984 32. Adzick NS, Harrison MR, Glick PL, et al: Fetal surgery in the primate III. Maternal outcome after fetal surgery. J Pediatr Snrg 21:477-480, 1986 33. Harrison, MR: Unpublished data. 34. Harrison MR, Golbus MS, Filly RA (eds): The Unborn Patient. Philadelphia, Grune & Stratton, 1984, pp 277-348 35. Glazer GM, Filly RA, Callen PW: The varied sonographic appearance of the urinary tract in the fetus and newborn with urethral obstruction. Radiology 144:563-568, 1982 36. Hobbins JC, Romero R, Gronnum P, et al: Antenatal diagnosis of renal anomalies with ultrasound I. Obstructive uropathy. Am J Obstet Gynecol 148:868-877, 1984
1120
CROMBLEHOLME ET AL
Discussion Stephen Dunn (Philadelphia): The lethal defect in these children is pulmonary hypoplasia. The fact that some of these children may not develop normal renal function doesn't seem to me to be such a bad thing. You are now salvaging children who have normal intelligence, with normal developmental capabilities otherwise, and allowing them to get to the point where they may undergo a renal transplant. I think all of the information that we have now would say that the prognosis with renal transplantation for a child developing renal failure in this age group is very good. I congratulate you on your results and I think that you should not stop performing this procedure even if you cannot achieve completely normal renal function. Correction of pulmonary hypoplasia is an important accomplishment in these children. M. Dimler (Phoenix): I would like to ask one question concerning the degree of dysplasia and its relationship to timing of intervention particularly with your non-human primate model. Do you have data that are helpful in sorting out the time course at which such severe dysplasia is present that one should not contemplate salvage? T. Weber (St Louis): Extracorporeal membrane oxygenation has redefined lethal pulmonary hypoplasia. In our series of 130 newborns treated with ECMO, we have two babies born with severe oligohydramnios, bell-shaped chest, and pulmonary hypoplasia; they were treated with ECMO and both have survived. I think at least some of the pathophysiology in these newborns is pulmonary hypertension rather than insufficient pulmonary parenchyma. We need to look at aggressive means of support for these babies, and they should still be considered candidates for ECMO. J. Noseworthy (Oncinnati): Certainly the San Francisco group continues to push us and challenge us to think about new methods of dealing with diseased lungs and kidneys. Would you comment about patientSelection despite the fact that you want that dealt with a little bit later on? I think it is an important aspect of what you have presented. Could you give us some general outlines? J. Langer (response): Dr Dunn, we would agree wholeheartedly with your comments. The field of transplantation is still in its infancy, and we are sure that further advances will allow children with chronic renal failure to have better qualities of life. Dr Dimler, it is always difficult to apply an animal model to the human situation, but certainly there seems to be a very strong relationship between the amount of time that the urinary tract is obstructed and the degree of renal damage. We do not know how many
weeks' gestation that translates to in the human fetus, and this underlines the importance of developing prognostic criteria that we can trust. One of these may be the gestational age at which hydronephrosis or oligohydramnios occurs. Dr Noseworthy, I think that there are two issues to deal with here. One is pulmonary outcome and the other is renal outcome, and when we look at prognostic criteria and selection, we have to think of them separately. It is now clear to us that the presence of early oligohydramnios is a very bad prognostic sign for pulmonary hypoplasia. The fetus who has oligohydramnios without evidence of longstanding pre-existing oligohydramnios or in particular the fetus who has hydronephrosis and then develops oligohydramnios, is the kind of patient at risk for pulmonary hypoplasia and should have in utero intervention. The prognostic criteria for the ultimate development of chronic renal failure is a far more difficult question. The things that we have looked at are fetal urinary electrolytes and osmolarity, ultrasound evidence of dysplasia in the kidneys, as well as the presence of oligohydramnios. We are well on our way to determining how to interpret these data, but we still have a long way to go. T. Bartholomew (San Antonio, TX): You have not mentioned anything about the complications of your intervention. Those are real; you pointed out one abortion from chorioamnitis, and also you haven't pointed out the accuracy of your diagnoses. These patients may have prune belly syndrome; in some of your drawings, in fact used on your slides, a patient is depicted with prune belly syndrome with an assumption that the intravesicle obstruction in some way caused the prue belly syndrome. I don't think that can be accepted. Do you have any comment on the accuracy of the diagnoses that have been involved and the complications of your intervention? T.M. Crombleholme (response): Dr Bartholomew, in terms of the complications, three vesicoamniotic catheter shunts were complicated by chorioamnioitis. These occurred early in our experience, prior to the use of prophylactic antibiotics. These have been no maternal complications as a result of their placement. In open fetal surgery, there have been two complications: pseudomembranous colitis and a single instance of amniotic fluid leak due to a technical problem that has been rectified and has not recurred. Stephen Dunn (Philadelphia): How do you choose either the Pigtail Harrison catheter or an open fetal procedure? How do your criteria assist you in making those decisions? The second comment I had was that I think your data on urine electrolytes is interesting, but
OPEN FETAL SURGERY FOR CONGENITAL HYDRONEPHROSIS
in using the electrolyte data as one of your criteria for separating good from poor prognosis groups, you expect them to have a significant difference in the final analysis in the first place. T.M. Crombleholme (response): Dr Dunn, you raise a crucial issue and one with which we continually struggle. We have not yet refined the criteria sufficiently to distinguish those patients in the poor prognostic group who will survive but go on to have renal failure from those with no hope of survival. S. Kramer (Rochester, NY): You have shown nicely the difference between the good and poor prognostic groups. You have not convinced me yet, though, that the good prognostic group would not do just as well if you left them alone and then did a vesicostomy or ureterostomy at birth. In the six patients you showed that did well with intervention, don't you think perhaps they would have done just as well if they were delivered at term and then had vesicostomies at that time? We have had that experience and there is a report in the Journal of Urology (May issue) of a patient with
1121
oligohydramnios and bilateral hydronephrosis who was delivered, had a vesicostomy, and now at 2 years of age, has normal pulmonary and normal renal function. T.M. Crombleholme (response): Dr Kramer, with regard to the issue of whether or not the good prognosis patients would have done well without intervention, six of the eight patients who survived had oligohydramnios, and would not have survived without restoration of amniotic fluid. Survival of patients with oligohydramnios diagnosed early remains anecdotal. T.M. Crombleholme (closing): The decision to use a vesicoamniotic shunt as opposed to an open procedure depends largely upon the time at which the patient presents. The functional life span of a vesicoamniotic shunt is less than 2 weeks. A patient who presents early in gestation (at 20 to 24 weeks) would require multiple catheter placements which are not without complications. In that instance, we would recommend open fetal surgery.