- Email: [email protected]
Vesicoamniotic shunt for complete urinary tract obstruction is partially effective
Journal of Pediatric Surgery (2006) 41, 394 – 402
www.elsevier.com/locate/jpedsurg
Vesicoamniotic shunt for complete urinary tract obstruction is partially effective Hiroaki Kitagawa a,*, Kevin C. Pringleb, Junki Koikec, Jane Zuccollob, Yasuji Seki a, Munechika Wakisakaa, Yuriko Satoa, Hideaki Satoa, Hideki Nagaea, Koonosuke Nakadaa a
Division of Pediatric Surgery, St. Marianna University School of Medicine, Kawasaki 216-8511, Japan Department of Surgery and Anaesthesia and Obstetrics and Gynaecology, Wellington School of Medicine, Wellington, New Zealand c Department of Pathology, St. Marianna University School of Medicine, Kawasaki 216-8511, Japan b
Index words: Obstructive uropathy; Fetal therapy; Multicystic dysplastic kidney; Vesicoamniotic shunt; Oligonephronia; Vesicoamniotic shunt; Bladder function
Abstract Purpose: The long-term outcome for children after antenatal intervention for obstructive uropathies is disappointing. We reported that renal dysplastic changes are well established 3 weeks after obstruction in a fetal lamb model. We used this model to explore renal development and bladder function after fetal intervention. Methods: We created an obstructive uropathy in fetal lambs at 60 days gestation by ligating the urethra and urachus. A vesicostomy (female) or urethrostomy (male) were performed 21 days later. The fetuses were killed at term (145 days) and bladder volume and compliance were measured. The urinary tract was processed for histologic examination. Results: Twenty two fetuses were shunted. Nine were miscarried or were still-born. Thirteen survived, and 11 had a successful shunt with a small bladder (8 F 5 mL) compared with controls (71 F 19 mL) ( P b .05). Shunted bladders had poor compliance. Histologically, they had thickened submucosal connective tissue with hypertrophied muscle. Histology of the renal tissue demonstrated relatively well-preserved renal architecture with reduced nephron mass (oligonephronia) in 2 lambs and multicystic dysplastic change in 3. Six (55%) had normal nephrogenesis. Conclusions: In our model, shunt operations after obstructive uropathy fail to preserve bladder function. Shunting ameliorated the development of cystic dysplasia, but half of the lambs had oligonephronia or multicystic dysplastic kidney. These might develop renal failure later in life. D 2006 Elsevier Inc. All rights reserved.
Presented at the 52nd Annual Congress of British Association of Paediatric Surgeons, Dublin, Ireland, July 12-15, 2005. 4 Corresponding author. E-mail address: [email protected] (H. Kitagawa ). 0022-3468/$ – see front matter D 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.jpedsurg.2005.11.035
Fetal surgery for obstructive uropathy started in humans during the 1980s at the University of California, San Francisco [1,2]. The developmental pathophysiology of several potentially lethal correctable lesions was defined in animal models [3,4]. Over the ensuing 20 years, this investment in basic and clinical research has benefited an
Vesicoamniotic shunt for urinary tract obstruction increasing number of fetal patients and has spurred continued research into the normal and abnormal development of the fetus and the implications of this type of therapy [5,6]. The clinical outcomes of antenatally diagnosed obstructive uropathy are now being reported, but the long-term results are not good as had been expected [7,8]. Coplen [9] reviewed the 5 largest series of successful fetal interventions. Although the fetal interventions were technically feasible, the overall survival was only 47% with significant fetal and maternal risks during both open fetal surgery and catheter placement procedures. This study suggests that minimally invasive fetal intervention did not improve outcome, despite the presence of favorable ultrasound findings and urinary electrolytes. Freedman et al [10] reported 55 consecutive patients with obstructive uropathy diagnosed by ultrasonography in 1996. Fetuses were classified as having good or bad prognosis based on fetal urinalysis. Of the 33 patients with a good prognosis, 22 underwent vesicoamniotic (V-A) shunting, and 64% survived compared with 45% survival of the nonshunted fetuses. As a comparison, 6 of 22 fetuses in the poor prognosis group underwent shunting and 50% survived [10]. Holmes et al [11] emphasized that it is important not to give families unrealistic expectations that fetal surgery is the cure for obstructive uropathy or that the child will not need extensive treatment after birth and long-term follow-up. Fetal treatment of obstructive uropathy now includes open fetal surgery, percutaneous V-A shunt placement, and fetoscopic procedures and is performed only for carefully selected patients. These changes in surgical technique have not changed the physiology of the bladder after shunting. Therefore, we attempted to create a model similar to posterior urethral valves in fetal lambs and then bypass the obstruction 3 weeks later to see whether it is possible to preserve normal renal and bladder development with appropriately timed fetal surgery [12]. The initial experimental human fetal surgery for obstructive uropathy was based on sound research data but unfortunately, the long-term follow-up results in humans suggests that only approximately 50% will be long-term survivors after V-A shunt procedures [7,9,11]. These results are nowhere near what would have been expected, based on the data obtained by previous investigators using the fetal lamb model [12]. For this study, we created a model similar to the posterior urethral valve (PUV) at 60 days in the fetal lamb and decided (based on our experimental studies [13]) that the optimal timing for fetal shunting would be 3 weeks after the obstruction was created. In this study, we attempted to set up a model that would enable us to compare the results of a shunting procedure that provided free egress of urine from the bladder with little resistance with a shunt that would allow the bladder to continue to develop in a situation where the bladder was able to continue with a filling/emptying cycle, with some resistance to outflow.
395
1. Materials and methods After approval was obtained from the Animal Ethics Committee of Wellington School of Medicine and Health Sciences, timed gestation ewes (60 days gestation) were transported from the farm 24 to 48 hours before operation. They were examined by ultrasound to confirm the pregnancy and avoid unnecessary operations. The ewes were initially induced by gas induction, but at the suggestion of the Animal Ethics Committee, we have changed to using intravenous thiopentone for the induction. The ewes were then intubated, once sufficiently anesthetized and anesthesia was maintained using nitrous oxide oxygen and halothane, as previously described [14]. The uterus was then delivered through a left flank incision. The lambs then underwent urethral and urachal ligation for male lambs, and female lambs had their bladder neck and urachus ligated using fine Silastic tubing (Down Corning, Midland, Mich) of 0.025 inch (0.625 mm) [6]. After the operation, the fetuses were returned to the uterus, and the ewe’s abdomen was closed. The ewes were then transported back to the farm once they had recovered from the anesthetic. Three weeks later, the ewes were transported back from the farm. They were again examined ultrasonographically to confirm the viability of the fetuses, and the dilated bladder or renal pelvis was identified. Under general anesthesia, the uterus was reopened using the previous incision, and the fetuses were exposed. In female fetuses a lower abdominal incision was made, and a small hole was made in the bladder wall to approximate the placement of a shunt but without the risk of displacement of the shunt. The bladder wall and skin were sutured to each other using 5-0 Biosyn (U.S. Surgical, Tyco Healthcare, Norwalk, CT; CV 22, GM 875, monofilament glycomer) to prevent closure of the vesicostomy. In males, the dilated urethra was opened through the perineum, creating a urethrostomy. A short segment of Silastic tubing (Down Corning, 2-mm catheter) was placed through the urethrostomy into the bladder as a stent and sutured to the perineum using a 4-0 Ethibond suture (Ethicon, Somerville, NJ) [15]. This was done in the hope that as the fetuses grew, the proximal end of the shunt would drop below the external sphincter, allowing the bladder to return to the normal cycle of filling and emptying. The fetuses were returned to the uterus and the uterus and ewe’s abdominal wall were closed, as previously outlined [6,14]. At the time of the initial operation, some of the lambs were randomly selected to be used as normal, nonoperated controls. These lambs were invariably twins with the other twin having been subjected to an obstructing operation. Similarly, at the time of the shunting procedure, some of the lambs were randomly selected to be left as a nonshunted obstructed uropathy model. Again, these lambs were usually twins, with the other twin receiving a shunting procedure. The fetuses were delivered at term (145 days) by caesarean section, and bladder compliance was measured
396
H. Kitagawa et al.
Fig. 1 The system used to measure the pressure-volume curve. The Pressure Amplifier (model PA-100) was connected to a 3F catheter type of pressure transducer (model AS301, Star Medical Inc, Tokyo, Japan), and this catheter was inserted into the bladder.
before the lambs were killed using pentobarbital injected into the umbilical vein. The initial bladder volume was measured by aspirating the urine through the bladder wall. We used a monitor system using a Pressure Amplifier
(model PA-100) connected to a 3F catheter pressure transducer (model AS301, Star Medical Inc, Japan; Web site: http://www.starmedical.co.jp) through an extension cable from the bladder to produce the pressure-volume
Fig. 2 A-D, In normal-term lambs, the right kidneys weighed 16.54 F 1.80 g (n = 5), and the left kidneys weighed 16.52 F 2.29 g (n = 5) (A). In the oligonephronia type of shunted kidneys, the right kidneys weighed 3 g, and left kidneys weighed 4.6 g (B). The renal cortex is then compared with normal-term lamb kidneys. Glomeruli appear to be decreased in number, but the uriniferous tubules are well developed (C, original magnification 20; D, original magnification 100).
Vesicoamniotic shunt for urinary tract obstruction
397
Fig. 3 A-C, In the cystic dysplastic type of shunted kidneys, the kidneys are larger than those of normal-term lamb kidneys, but the cortex contains multiple cysts (A). Both grossly and on microscopy, these findings are similar to our nonshunted 60-day obstructive uropathy model, except that in the shunted kidneys, there is a collar of fibrous tissue in the walls of the cysts (black arrow) that is not seen in the nonshunted cystic dysplastic kidneys. In the shunted kidneys, the cysts are lined with proximal tubular epithelium (B, original magnification 20; C, original magnification 100).
curve. The 3F catheter pressure transducer was inserted into the bladder through the vesicostomy or urethrostomy, and a purse string suture was applied around the stoma to prevent
leakage around the catheter (Fig. 1). Normal saline was injected into the bladder (10 mL/min), and the bladder pressure curve was recorded by the monitor.
Fig. 4 A, C, Six lambs had kidneys with mild pelvic dilatation (A) and normal renal histology (B, original magnification 20; C, original magnification 40).
398 The lambs were killed by injecting 10 mL of Barbital into the umbilical vein. The lamb’s urinary system was then removed en bloc, and the body weight, crown-to-rump length, kidney weights, and bladder capacity were measured.
H. Kitagawa et al. The kidneys were fixed in formal saline for histologic examination. Data from the shunted lambs were compared with data from normal control lambs and nonshunted obstructed lambs and analyzed by Student’s t test or
Fig. 5 A-D, The pressure-volume curve of a normal term control lamb (A). The pressure began to increase between 3 and 4 minutes (30 - 40 mL) after commencing the injection of normal saline into the bladder. At 5 minutes after beginning the infusion, the pressure remained at around 40 mm Hg as the size of bladder gradually increased. A pressure-volume curve of a V-A shunted bladder (B, C, and D). B demonstrates that the pressure in this bladder began to rise very steeply after the infusion of only 8 mL of saline, reaching 60 mm Hg within 30 seconds of commencing the infusion. C, This bladder volume was approximately 20 mL. The pressure started to rise at 1 minute after commencing the infusion. The pressure reached to 100 mm Hg within 2 minutes (20 mL). D, the bladder volume was 8 mL. Ninety seconds after commencing the infusion of normal saline into the bladder, the bladder pressure rose to more than 160 mm Hg.
Vesicoamniotic shunt for urinary tract obstruction Table 1 Bladder compliance and volume curve of control and shunt group Bladder volume at sacrifice (mL)
Volume to initial pressure rise (s)
Final volume (mL)
Final pressure (mm Hg)
Control 71 F 19 180 -240 40 - 50 40 (n = 2) Shunt (n = 7) 9.6 F 5.5 70 F 19 10.6 F 5.5 100 F 43.2
Welch’s test, as appropriate. No true sham-operated lambs were used in these experiments. At the time these experiments were designed, we were approaching the technical limitations of these models, and it was decided that because survival of the experimental lambs was expected to be only in the region of 50%, a model using a shunted nonobstructed bladder would have compromised our ability to have enough of the experimental lambs surviving to sacrifice to enable appropriate conclusions to be drawn.
2. Results We operated on 31 fetuses to create a model of obstructive uropathy at 60 days of gestation. Two of them died after creation of obstruction. Five of the lambs did not undergo shunting and were kept until term as nonshunted controls. One ewe with twin fetuses became very ill a few days after the initial operation, and she was euthanized.
399 Twenty two fetuses were shunted at 81 days (3 weeks after the initial operation). Nine fetuses were miscarried or were still-born after shunting. Thirteen survived, and 11 lambs had a successful shunt with a small bladder volume (8 F 5 mL) (bshuntedQ lambs). Body weight and crown-rump length were compared between the shunt model and normal nonoperated term lambs and nonshunted obstructed term lambs. The body weight was 4870.2 F 626.4 g for normal control lambs (n = 5), 4236 F 887 g in the shunt model (n = 11), and 4699.6 F 483.3 g in nonshunted obstructed lambs (n = 4). The crown-rump length was 53 F 6.8 cm for normal control lambs, 48 F 3.6 cm in the shunt model, and 48.6 F 2.3 in nonshunt controls. There was no statistical difference between these groups using unpaired Student’s t test. There were no signs of intrauterine growth retardation after the shunt operation. Macroscopically, in the shunted lambs, there were 3 different types of fetal kidney. Two fetuses had very small kidneys compared with control kidneys (Fig. 2A and B). In these lambs, the right kidneys weighed 9.0 g and 3.0 g, and left kidneys weighed 9.0 g and 4.6 g. In normalterm lambs, the right kidneys weighed 16.54 F 1.80 g (n = 5), and the left kidneys weighed 16.52 F 2.29 g (n = 5). Microscopically, these small kidneys demonstrated a thin renal cortex compared with normal-term lamb kidneys. Glomeruli were decreased in number, but the proximal and distal tubules were well developed. We have named this small-sized kidney with an apparent reduction in the number of glomeruli the boligonephroniaQ type of obstructed kidney (Fig. 2C and D). Three lambs had cystic dysplastic changes
Fig. 6 A and B (Masson’s trichrome stain) demonstrate the normal thin bladder wall. Submucosal fibrous connective tissue is visible, but no intramuscular fibrosis is recognized. On immunohistochemistry staining fora-SMA, the muscular layer, but not the submucosal, stain positively for SMA (C).
400 even after the shunt operation. In these lambs, the kidney weighed 67.8 F 50 g (n = 6 kidneys). Macroscopically, there were large cysts in the cortex (Fig. 3A). These findings are similar to our nonshunted 60-day obstructive uropathy model, but the difference between the shunt model and nonshunt obstructive uropathy model is that in the shunted lambs, on histologic examination, the walls of the cysts in kidneys had a fibrous collar and a lining of proximal tubular epithelium (Fig. 3B and C). Six lambs had normal renal histology with macroscopic pelvic dilatation (Fig. 4A-C). The bladder compliance was measured using pressure volume curve in 7 shunted cases and 2 control fetuses. The control bladder pressure curve started to rise from 4 to 5 minutes (40-50 mL of saline injected) after the infusion of normal saline commenced (Fig. 5A). In the V-A shunted bladders, in lamb 1, it took only 30 seconds (approximately 5 mL) for the pressure to increase. The pressure curve was very steep, and saline started to leak from the vesicostomy as soon as the pressure began to rise quickly (Fig. 5B). The third V-A shunted lamb had 20 mL of bladder capacity. In this lamb, the pressure began to rise from 1 minute after commencing the injection, and by 2 minutes, the bladder pressure rose to more than 100 mm Hg (Fig. 5C). The fourth lamb also had a poorly compliant bladder and the pressure curve rose to more than 160 mm Hg within 90 seconds (Fig. 5D). Most of the shunted bladders showed a steep rise in the pressure curve within 30 seconds to 1 minute, and all the curves were very steep compared with the controls. All
H. Kitagawa et al. of the shunted bladders started leaking after injecting 4 to 20mL of saline (Table 1). Histologically, the normal bladder wall was thin, compared with the bladder wall seen in the shunt model. The submucosal fibrous connective tissue is visible, but no intramuscular fibrosis is recognized. On immunohistochemical staining for smooth muscle actin (SMA), the muscular layer stained positive but not the submucosal area (Fig. 6). In the shunt model, there was marked bladder wall thickening with dense fibrosis extending from the submucosal layer into the muscular layers. On immunohistochemical staining for SMA, myofibroblastic cell proliferation was recognized in the submucosal layer. The muscular layer was also thickened, but the muscular bands were fragmented (Fig. 7).
3. Discussion Our previous studies have shown that there is a cystic dilatation that develops in the renal cortex, and this is mainly because of dilatation of the proximal tubules, which started to dilate at 48 hours after the obstruction and continued to develop over the first 7 days after the obstruction [15]. This study was designed to attempt to determine when this dilatation becomes irreversible. After the creation of a lower urinary tract obstruction, cystic dilatation of the ampullae in the nephrogenic zone
Fig. 7 In the bladder of a shunted lamb (Masson’s trichrome stain) (A and B), there is marked thickening of the bladder wall with fibrosis extending from the submucosal layer into the intramuscular region. On immunohistochemistry staining for SMA (C), myofibroblastic cell proliferation is recognized in the submucosal layer. The muscular layer is also thickened, but the muscular bands are fragmented by thick bands of collagen.
Vesicoamniotic shunt for urinary tract obstruction suppresses nephrogenesis [13]. This hypothesis may explain the relation between the timing of the obstruction and its effect on the developing kidney. Our previous study demonstrated that when we created an obstructive uropathy at 60 days, the nephrogenic zone is still intact 3 weeks after the obstruction [13]. If the nephrogenic zone is still intact, then our hypothesis was that the kidney would have a chance to produce new nephrons if the obstruction was relieved. In our model of obstructive uropathy created at 60 days in fetal lambs, we showed that in a series of nonshunted obstructed term lambs, half of the lambs had small kidneys with small numbers of dilated tubules in the cortex. However, 25% of the lambs in our previous study had bmacrocysticQ kidneys with prominent cystic dilatation of tubules in the mid to deep cortex [13]. Creating identical obstructions at identical times in renal development produced markedly different phenotypes of renal dysplasia [16]. We hypothesized that if a V-A shunt was able to release the pressure in the proximal tubules, the cystic dilatation will resolve, and new nephrons will be able to develop from the nephrogenic zone. We speculated that in the bmicrocysticQ kidneys, the cysts may also disappear, and new nephrons will continue to be produced until term. In human kidneys, the rate of development of glomerular populations during fetal life means that the total numbers of glomeruli rapidly increase before 24 weeks of gestation, and from 24 to 40 weeks, few new glomeruli are produced [17,18]. Our oligonephronia model had only small numbers of nephrons. This is related to this growth curve in the glomerular population. In the bmacrocysticQ kidneys in the shunted lambs, the cysts persist, and the dilated proximal tubules developed a cuff of connective tissue. Why this occurred is not clear but these 2 different types of kidney are somehow related to the dysplastic kidneys we reported in our previous paper in a nonshunted model [13]. Our results in this study suggest that a shunt operation for obstructive uropathy 3 weeks after an obstruction is produced in fetal lambs at 60 days of gestation is partially effective. The human equivalent of the timing of our shunting procedures is approximately 15 to 17 weeks of gestation. Most human shunting procedures are carried out at 20 to 24 weeks of gestation [19]. In human fetuses, the shunt-related complication rate is high if the shunts are performed at an early stage in gestation [8,20]. However, if optimal renal and bladder function are to be preserved, we are convinced that it is vital to ensure the development of normal bladder physiology by maintaining the normal voiding cycle. If an infant survives after the successful in utero placement of a V-A shunt, the next step is to preserve bladder function after shunting. This may be more difficult to assess than the assessment of the preservation of renal function. We reported the effect of early vesicostomy in obstructive uropathy on bladder development and found that after early shunting, the bladder wall was thickened with increased fibrosis in the submucosal layer [21]. These bladder abnormalities are probably the result of the initial
401 complete obstruction causing secondary changes to the development of the muscle walls in the bladder, resulting in an increased deposition of connective tissue [21]. Bladder hypertrophy is a common finding among patients with posterior urethral valves [22,23]. The hypertrophy is so marked that bladder compliance is significantly reduced. Our animal model demonstrated that the V-A shunted bladder can contain only 8.8 F 4.7 mL, and this is approximately one tenth of the capacity of the bladder in control animals. The bladder pressure-volume curve demonstrated that the V-A shunted model had extremely poor bladder compliance. In our nonshuntedterm lambs with obstructive uropathy, the severely dilated bladder demonstrated increased connective tissue deposition between the detrusor muscle bundles. After a V-A shunt for obstructive uropathy, the fetuses urinate through the shunt, and this procedure removes the normal bladder physiological cycle of filling and emptying against a slight resistance. The bladder capacity is very small because the shunt continuously drained the urine into the amniotic fluid. The problem is not only the size of the bladder but also the severely reduced bladder compliance. We used 2 different types of bypass procedure in this series of experiments. The vesicostomy was a surrogate for a V-A shunt but avoided the possibility of the shunt being dislodged. The placement of the perineal urethrostomy in the male lambs was an attempt to place a catheter that would descend below the external sphincter as the lamb grew, allowing the bladder to return to a normal filling and emptying cycle. Unfortunately, in all of the lambs, the catheter placed was too long and never descended below the external sphincter. This meant that it functioned in exactly the same way as the vesicostomy. In long-term follow-up of 14 children, Freedman et al [24] reported that 8 children void spontaneously, 4 children catheterise in addition to voiding, and 2 children catheterised only. Of the 4 children with posterior urethral valves, 3 have undergone bladder augmentation. Holmes et al [11] reported that 5 of the 8 survivors had some type of postnatal urinary diversion or augmentation. Our fetal lamb model is very similar to the posterior urethral valve patients, and this group might need bladder augmentation a few years after birth to increase bladder volume. From the histologic point of view, the SMA is transiently expressed in the myocardium and skeletal muscle during the development of the embryo and is highly restricted in adult animals to smooth muscle cells [25,26]. However, it is also expressed in myofibroblasts during wound healing and in a very limited number of adult tissues [27]. Little is known regarding the molecular mechanisms that control the activation of SMA in myofibroblasts in granulation tissue. On Masson’s trichrome stain, massive collagen deposition is observed in the bladder wall. Our immunohistochemical results demonstrated that the submucosal area represents proliferation of SMA-positive fibroblastic cells (bmyofibroblastsQ) in the walls of the shunted bladders, and
402 this results in the observed fibrosis in the bladder wall. Myofibroblasts are specialized contractile fibroblasts that are critical in wound closure and tissue contracture. From our animal data, the muscle hypertrophy and submucosal fibrosis are the key factors that result in the striking reduction in bladder compliance we have observed. Once severely dilated, the fetal bladder had increased connective tissue in the submucosal layer, extending into the intramuscular region. Muscle thickness was markedly increased in obstructed fetal bladders as the lambs approached term. If the muscle thickness increases with fetal age, an early shunt may preserve normal bladder compliance. Early shunting or the use of a pressure controlled shunt tube is the next phase in our studies.
Acknowledgment This study was supported by the Japanese Society for Grant aid for Scientific Research (C). Suture materials (Biosyn) were supplied by Tyco, Healthcare, Tokyo, Japan. These experiments would not have been possible without the skill and dedication of Doug Jensen of the Wallaceville Animal Research Facility. He consistently produces animals of timed gestation with the time of mating recorded to within 24 hours.
References [1] Golbus MS, Harrison MR, Filly RA, et al. In utero treatment of urinary tract obstruction. Am J Obstet Gynecol 1982;142:383 - 8. [2] Harrison MR, Golbus MS, Filly RA, et al. Fetal surgery for congenital hydronephrosis. N Engl J Med 1982;306:591 - 3. [3] Tanagho EA. Surgically induced partial urinary obstruction in the fetal lamb III. Ureteral obstruction. Invest Urol 1972;10:35 - 52. [4] Harrison MR, Nakayama DK, Noall R, et al. Correction of congenital hydronephrosis in utero. II. Decompression reverses the effect of obstruction on the fetal lung and urinary tract. J Pediatr Surg 1982; 17:965 - 74. [5] Glick PL, Harrison MR, Adzic NS, et al. Correction of congenital hydronephrosis in utero. IV. In utero decompression prevents renal dysplasia. J Pediatr Surg 1984;19:649 - 57. [6] Kitagawa H, Pringle KC, Zuccollo J, et al. The pathogenesis of dysplastic kidney in a urinary tract obstruction in the female fetal lamb. J Pediatr Surg 1999;34:1678 - 83. [7] Freedman AL, Johnson MP, Smith CA, et al. Long-term outcome in children after antenatal intervention for obstructive uropathy. Lancet 1999;354:374 - 7.
H. Kitagawa et al. [8] Freedman AL, Johnson MP, Gonzalez R. Fetal therapy for obstructive uropathy: past, present, . . .future? Pediatr Nephrol 2000;14:167 - 76. [9] Coplen DE. Prenatal intervention for hydronephrosis. J Urol 1997; 157:2270 - 7. [10] Freedman AL, Bukowski TP, Smith CA, et al. Fetal therapy for obstructive uropathy: diagnosis specific outcome. J Urol 1996;156:720 - 3. [11] Holmes N, Harrison MR, Baskin LS. Fetal surgery for posterior urethral valves: long-term postnatal outcome. Pediatrics 2001;108:E7. [12] Harrison MR, Filly RA. The fetus with obstructive uropathy: pathophysiology, natural history, selection, and treatment. In: Harrison MR, Golbus MS, Filly RA editors. The unborn patient: prenatal diagnosis and treatment. Philadelphia7 WB Saunders; 1991. p. 328 - 93. [13] Kitagawa H, Pringle KC, Koike J, et al. Optimal timing of perinatal treatment of obstructive uropathy in the fetal lamb. J Pediatr Surg 2003;38:1785 - 9. [14] Pringle KC. Fetal lamb and fetal lamb lung growth following creation and repair of a diaphragmatic hernia. In: Nathanielsz PW, editor. Animal models in fetal medicine. Ithaca (NY)7 Perinatology Press;1984. p. 109 - 48 [chapter 6]. [15] Kitagawa H, Pringle KC, Koike J, et al. The early effects of urinary tract obstruction on glomerulogenesis. J Pediatr Surg 2004;39:1845 - 8. [16] Kitagawa H, Pringle KC, Koike J, et al. Different phenotypes of dysplastic kidney in obstructive uropathy in fetal lambs. J Pediatr Surg 2001;36:1698 - 703. [17] Osathanondh V, Potter EL. Pathogenesis of polycystic kidneys. Survey of results of microdissection. Arch Pathol 1964;77:510 - 2. [18] Potter EL. Normal and abnormal development of the kidney. Chicago7 Year Book Medical Publishers, Inc; 1972. p. 15 - 21. [19] Clark TJ, Martin WL, Divakaran TG, et al. Prenatal bladder drainage in the management of fetal lower urinary tract obstruction: a systematic review and meta-analysis. Am J Obstet Gynecol 2003;102:367 - 82. [20] Jouannic JM, Hyett JA, Pandya PP, et al. Perinatal outcome in fetuses with megacystis in the first half of pregnancy. Prenat Diagn 2003; 23:340 - 4. [21] Sato Y, Kitagawa H, Pringle KC, et al. Effect of early vesicostomy in obstructive uropathy on bladder development. J Pediatr Surg 2004; 39:1849 - 52. [22] Bauer SB, Dieppa RA, Labib KK, et al. The bladder in boys with posterior urethral valve: a urodynamic assessment. J Urol 1979;121: 769 - 73. [23] Peters CA, Bolkier M, Bauer SB, et al. The urodynamic consequences of posterior urethral valves. J Urol 1990;144:122 - 6. [24] Freedman AL, Johnson MP, Smith CA, et al. Long-term outcome in children following antenatal intervention for obstructive uropathies. Pediatrics 1997;100:578. [25] Tomasek JJ, McRae J, Owens GK, et al. Regulation of alpha-smooth muscle actin expression in granulation tissue myofibroblasts is dependent on the intronic CarG element and the transforming growth factor-beta1 control element. Am J Pathol 2005;166:1343 - 51. [26] Hinz B, Gabbiani G. Mechanisms of force generation and transmission by myofibroblasts. Curr Opin Biotechnol 2003;14:538 - 46. [27] Darby I, Skalli O, Gabbiani G. Alpha-smooth muscle actin is transiently expressed by myofibroblasts during experimental wound healing. Lab Invest 1990;63:21 - 9.
www.elsevier.com/locate/jpedsurg
Vesicoamniotic shunt for complete urinary tract obstruction is partially effective Hiroaki Kitagawa a,*, Kevin C. Pringleb, Junki Koikec, Jane Zuccollob, Yasuji Seki a, Munechika Wakisakaa, Yuriko Satoa, Hideaki Satoa, Hideki Nagaea, Koonosuke Nakadaa a
Division of Pediatric Surgery, St. Marianna University School of Medicine, Kawasaki 216-8511, Japan Department of Surgery and Anaesthesia and Obstetrics and Gynaecology, Wellington School of Medicine, Wellington, New Zealand c Department of Pathology, St. Marianna University School of Medicine, Kawasaki 216-8511, Japan b
Index words: Obstructive uropathy; Fetal therapy; Multicystic dysplastic kidney; Vesicoamniotic shunt; Oligonephronia; Vesicoamniotic shunt; Bladder function
Abstract Purpose: The long-term outcome for children after antenatal intervention for obstructive uropathies is disappointing. We reported that renal dysplastic changes are well established 3 weeks after obstruction in a fetal lamb model. We used this model to explore renal development and bladder function after fetal intervention. Methods: We created an obstructive uropathy in fetal lambs at 60 days gestation by ligating the urethra and urachus. A vesicostomy (female) or urethrostomy (male) were performed 21 days later. The fetuses were killed at term (145 days) and bladder volume and compliance were measured. The urinary tract was processed for histologic examination. Results: Twenty two fetuses were shunted. Nine were miscarried or were still-born. Thirteen survived, and 11 had a successful shunt with a small bladder (8 F 5 mL) compared with controls (71 F 19 mL) ( P b .05). Shunted bladders had poor compliance. Histologically, they had thickened submucosal connective tissue with hypertrophied muscle. Histology of the renal tissue demonstrated relatively well-preserved renal architecture with reduced nephron mass (oligonephronia) in 2 lambs and multicystic dysplastic change in 3. Six (55%) had normal nephrogenesis. Conclusions: In our model, shunt operations after obstructive uropathy fail to preserve bladder function. Shunting ameliorated the development of cystic dysplasia, but half of the lambs had oligonephronia or multicystic dysplastic kidney. These might develop renal failure later in life. D 2006 Elsevier Inc. All rights reserved.
Presented at the 52nd Annual Congress of British Association of Paediatric Surgeons, Dublin, Ireland, July 12-15, 2005. 4 Corresponding author. E-mail address: [email protected] (H. Kitagawa ). 0022-3468/$ – see front matter D 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.jpedsurg.2005.11.035
Fetal surgery for obstructive uropathy started in humans during the 1980s at the University of California, San Francisco [1,2]. The developmental pathophysiology of several potentially lethal correctable lesions was defined in animal models [3,4]. Over the ensuing 20 years, this investment in basic and clinical research has benefited an
Vesicoamniotic shunt for urinary tract obstruction increasing number of fetal patients and has spurred continued research into the normal and abnormal development of the fetus and the implications of this type of therapy [5,6]. The clinical outcomes of antenatally diagnosed obstructive uropathy are now being reported, but the long-term results are not good as had been expected [7,8]. Coplen [9] reviewed the 5 largest series of successful fetal interventions. Although the fetal interventions were technically feasible, the overall survival was only 47% with significant fetal and maternal risks during both open fetal surgery and catheter placement procedures. This study suggests that minimally invasive fetal intervention did not improve outcome, despite the presence of favorable ultrasound findings and urinary electrolytes. Freedman et al [10] reported 55 consecutive patients with obstructive uropathy diagnosed by ultrasonography in 1996. Fetuses were classified as having good or bad prognosis based on fetal urinalysis. Of the 33 patients with a good prognosis, 22 underwent vesicoamniotic (V-A) shunting, and 64% survived compared with 45% survival of the nonshunted fetuses. As a comparison, 6 of 22 fetuses in the poor prognosis group underwent shunting and 50% survived [10]. Holmes et al [11] emphasized that it is important not to give families unrealistic expectations that fetal surgery is the cure for obstructive uropathy or that the child will not need extensive treatment after birth and long-term follow-up. Fetal treatment of obstructive uropathy now includes open fetal surgery, percutaneous V-A shunt placement, and fetoscopic procedures and is performed only for carefully selected patients. These changes in surgical technique have not changed the physiology of the bladder after shunting. Therefore, we attempted to create a model similar to posterior urethral valves in fetal lambs and then bypass the obstruction 3 weeks later to see whether it is possible to preserve normal renal and bladder development with appropriately timed fetal surgery [12]. The initial experimental human fetal surgery for obstructive uropathy was based on sound research data but unfortunately, the long-term follow-up results in humans suggests that only approximately 50% will be long-term survivors after V-A shunt procedures [7,9,11]. These results are nowhere near what would have been expected, based on the data obtained by previous investigators using the fetal lamb model [12]. For this study, we created a model similar to the posterior urethral valve (PUV) at 60 days in the fetal lamb and decided (based on our experimental studies [13]) that the optimal timing for fetal shunting would be 3 weeks after the obstruction was created. In this study, we attempted to set up a model that would enable us to compare the results of a shunting procedure that provided free egress of urine from the bladder with little resistance with a shunt that would allow the bladder to continue to develop in a situation where the bladder was able to continue with a filling/emptying cycle, with some resistance to outflow.
395
1. Materials and methods After approval was obtained from the Animal Ethics Committee of Wellington School of Medicine and Health Sciences, timed gestation ewes (60 days gestation) were transported from the farm 24 to 48 hours before operation. They were examined by ultrasound to confirm the pregnancy and avoid unnecessary operations. The ewes were initially induced by gas induction, but at the suggestion of the Animal Ethics Committee, we have changed to using intravenous thiopentone for the induction. The ewes were then intubated, once sufficiently anesthetized and anesthesia was maintained using nitrous oxide oxygen and halothane, as previously described [14]. The uterus was then delivered through a left flank incision. The lambs then underwent urethral and urachal ligation for male lambs, and female lambs had their bladder neck and urachus ligated using fine Silastic tubing (Down Corning, Midland, Mich) of 0.025 inch (0.625 mm) [6]. After the operation, the fetuses were returned to the uterus, and the ewe’s abdomen was closed. The ewes were then transported back to the farm once they had recovered from the anesthetic. Three weeks later, the ewes were transported back from the farm. They were again examined ultrasonographically to confirm the viability of the fetuses, and the dilated bladder or renal pelvis was identified. Under general anesthesia, the uterus was reopened using the previous incision, and the fetuses were exposed. In female fetuses a lower abdominal incision was made, and a small hole was made in the bladder wall to approximate the placement of a shunt but without the risk of displacement of the shunt. The bladder wall and skin were sutured to each other using 5-0 Biosyn (U.S. Surgical, Tyco Healthcare, Norwalk, CT; CV 22, GM 875, monofilament glycomer) to prevent closure of the vesicostomy. In males, the dilated urethra was opened through the perineum, creating a urethrostomy. A short segment of Silastic tubing (Down Corning, 2-mm catheter) was placed through the urethrostomy into the bladder as a stent and sutured to the perineum using a 4-0 Ethibond suture (Ethicon, Somerville, NJ) [15]. This was done in the hope that as the fetuses grew, the proximal end of the shunt would drop below the external sphincter, allowing the bladder to return to the normal cycle of filling and emptying. The fetuses were returned to the uterus and the uterus and ewe’s abdominal wall were closed, as previously outlined [6,14]. At the time of the initial operation, some of the lambs were randomly selected to be used as normal, nonoperated controls. These lambs were invariably twins with the other twin having been subjected to an obstructing operation. Similarly, at the time of the shunting procedure, some of the lambs were randomly selected to be left as a nonshunted obstructed uropathy model. Again, these lambs were usually twins, with the other twin receiving a shunting procedure. The fetuses were delivered at term (145 days) by caesarean section, and bladder compliance was measured
396
H. Kitagawa et al.
Fig. 1 The system used to measure the pressure-volume curve. The Pressure Amplifier (model PA-100) was connected to a 3F catheter type of pressure transducer (model AS301, Star Medical Inc, Tokyo, Japan), and this catheter was inserted into the bladder.
before the lambs were killed using pentobarbital injected into the umbilical vein. The initial bladder volume was measured by aspirating the urine through the bladder wall. We used a monitor system using a Pressure Amplifier
(model PA-100) connected to a 3F catheter pressure transducer (model AS301, Star Medical Inc, Japan; Web site: http://www.starmedical.co.jp) through an extension cable from the bladder to produce the pressure-volume
Fig. 2 A-D, In normal-term lambs, the right kidneys weighed 16.54 F 1.80 g (n = 5), and the left kidneys weighed 16.52 F 2.29 g (n = 5) (A). In the oligonephronia type of shunted kidneys, the right kidneys weighed 3 g, and left kidneys weighed 4.6 g (B). The renal cortex is then compared with normal-term lamb kidneys. Glomeruli appear to be decreased in number, but the uriniferous tubules are well developed (C, original magnification 20; D, original magnification 100).
Vesicoamniotic shunt for urinary tract obstruction
397
Fig. 3 A-C, In the cystic dysplastic type of shunted kidneys, the kidneys are larger than those of normal-term lamb kidneys, but the cortex contains multiple cysts (A). Both grossly and on microscopy, these findings are similar to our nonshunted 60-day obstructive uropathy model, except that in the shunted kidneys, there is a collar of fibrous tissue in the walls of the cysts (black arrow) that is not seen in the nonshunted cystic dysplastic kidneys. In the shunted kidneys, the cysts are lined with proximal tubular epithelium (B, original magnification 20; C, original magnification 100).
curve. The 3F catheter pressure transducer was inserted into the bladder through the vesicostomy or urethrostomy, and a purse string suture was applied around the stoma to prevent
leakage around the catheter (Fig. 1). Normal saline was injected into the bladder (10 mL/min), and the bladder pressure curve was recorded by the monitor.
Fig. 4 A, C, Six lambs had kidneys with mild pelvic dilatation (A) and normal renal histology (B, original magnification 20; C, original magnification 40).
398 The lambs were killed by injecting 10 mL of Barbital into the umbilical vein. The lamb’s urinary system was then removed en bloc, and the body weight, crown-to-rump length, kidney weights, and bladder capacity were measured.
H. Kitagawa et al. The kidneys were fixed in formal saline for histologic examination. Data from the shunted lambs were compared with data from normal control lambs and nonshunted obstructed lambs and analyzed by Student’s t test or
Fig. 5 A-D, The pressure-volume curve of a normal term control lamb (A). The pressure began to increase between 3 and 4 minutes (30 - 40 mL) after commencing the injection of normal saline into the bladder. At 5 minutes after beginning the infusion, the pressure remained at around 40 mm Hg as the size of bladder gradually increased. A pressure-volume curve of a V-A shunted bladder (B, C, and D). B demonstrates that the pressure in this bladder began to rise very steeply after the infusion of only 8 mL of saline, reaching 60 mm Hg within 30 seconds of commencing the infusion. C, This bladder volume was approximately 20 mL. The pressure started to rise at 1 minute after commencing the infusion. The pressure reached to 100 mm Hg within 2 minutes (20 mL). D, the bladder volume was 8 mL. Ninety seconds after commencing the infusion of normal saline into the bladder, the bladder pressure rose to more than 160 mm Hg.
Vesicoamniotic shunt for urinary tract obstruction Table 1 Bladder compliance and volume curve of control and shunt group Bladder volume at sacrifice (mL)
Volume to initial pressure rise (s)
Final volume (mL)
Final pressure (mm Hg)
Control 71 F 19 180 -240 40 - 50 40 (n = 2) Shunt (n = 7) 9.6 F 5.5 70 F 19 10.6 F 5.5 100 F 43.2
Welch’s test, as appropriate. No true sham-operated lambs were used in these experiments. At the time these experiments were designed, we were approaching the technical limitations of these models, and it was decided that because survival of the experimental lambs was expected to be only in the region of 50%, a model using a shunted nonobstructed bladder would have compromised our ability to have enough of the experimental lambs surviving to sacrifice to enable appropriate conclusions to be drawn.
2. Results We operated on 31 fetuses to create a model of obstructive uropathy at 60 days of gestation. Two of them died after creation of obstruction. Five of the lambs did not undergo shunting and were kept until term as nonshunted controls. One ewe with twin fetuses became very ill a few days after the initial operation, and she was euthanized.
399 Twenty two fetuses were shunted at 81 days (3 weeks after the initial operation). Nine fetuses were miscarried or were still-born after shunting. Thirteen survived, and 11 lambs had a successful shunt with a small bladder volume (8 F 5 mL) (bshuntedQ lambs). Body weight and crown-rump length were compared between the shunt model and normal nonoperated term lambs and nonshunted obstructed term lambs. The body weight was 4870.2 F 626.4 g for normal control lambs (n = 5), 4236 F 887 g in the shunt model (n = 11), and 4699.6 F 483.3 g in nonshunted obstructed lambs (n = 4). The crown-rump length was 53 F 6.8 cm for normal control lambs, 48 F 3.6 cm in the shunt model, and 48.6 F 2.3 in nonshunt controls. There was no statistical difference between these groups using unpaired Student’s t test. There were no signs of intrauterine growth retardation after the shunt operation. Macroscopically, in the shunted lambs, there were 3 different types of fetal kidney. Two fetuses had very small kidneys compared with control kidneys (Fig. 2A and B). In these lambs, the right kidneys weighed 9.0 g and 3.0 g, and left kidneys weighed 9.0 g and 4.6 g. In normalterm lambs, the right kidneys weighed 16.54 F 1.80 g (n = 5), and the left kidneys weighed 16.52 F 2.29 g (n = 5). Microscopically, these small kidneys demonstrated a thin renal cortex compared with normal-term lamb kidneys. Glomeruli were decreased in number, but the proximal and distal tubules were well developed. We have named this small-sized kidney with an apparent reduction in the number of glomeruli the boligonephroniaQ type of obstructed kidney (Fig. 2C and D). Three lambs had cystic dysplastic changes
Fig. 6 A and B (Masson’s trichrome stain) demonstrate the normal thin bladder wall. Submucosal fibrous connective tissue is visible, but no intramuscular fibrosis is recognized. On immunohistochemistry staining fora-SMA, the muscular layer, but not the submucosal, stain positively for SMA (C).
400 even after the shunt operation. In these lambs, the kidney weighed 67.8 F 50 g (n = 6 kidneys). Macroscopically, there were large cysts in the cortex (Fig. 3A). These findings are similar to our nonshunted 60-day obstructive uropathy model, but the difference between the shunt model and nonshunt obstructive uropathy model is that in the shunted lambs, on histologic examination, the walls of the cysts in kidneys had a fibrous collar and a lining of proximal tubular epithelium (Fig. 3B and C). Six lambs had normal renal histology with macroscopic pelvic dilatation (Fig. 4A-C). The bladder compliance was measured using pressure volume curve in 7 shunted cases and 2 control fetuses. The control bladder pressure curve started to rise from 4 to 5 minutes (40-50 mL of saline injected) after the infusion of normal saline commenced (Fig. 5A). In the V-A shunted bladders, in lamb 1, it took only 30 seconds (approximately 5 mL) for the pressure to increase. The pressure curve was very steep, and saline started to leak from the vesicostomy as soon as the pressure began to rise quickly (Fig. 5B). The third V-A shunted lamb had 20 mL of bladder capacity. In this lamb, the pressure began to rise from 1 minute after commencing the injection, and by 2 minutes, the bladder pressure rose to more than 100 mm Hg (Fig. 5C). The fourth lamb also had a poorly compliant bladder and the pressure curve rose to more than 160 mm Hg within 90 seconds (Fig. 5D). Most of the shunted bladders showed a steep rise in the pressure curve within 30 seconds to 1 minute, and all the curves were very steep compared with the controls. All
H. Kitagawa et al. of the shunted bladders started leaking after injecting 4 to 20mL of saline (Table 1). Histologically, the normal bladder wall was thin, compared with the bladder wall seen in the shunt model. The submucosal fibrous connective tissue is visible, but no intramuscular fibrosis is recognized. On immunohistochemical staining for smooth muscle actin (SMA), the muscular layer stained positive but not the submucosal area (Fig. 6). In the shunt model, there was marked bladder wall thickening with dense fibrosis extending from the submucosal layer into the muscular layers. On immunohistochemical staining for SMA, myofibroblastic cell proliferation was recognized in the submucosal layer. The muscular layer was also thickened, but the muscular bands were fragmented (Fig. 7).
3. Discussion Our previous studies have shown that there is a cystic dilatation that develops in the renal cortex, and this is mainly because of dilatation of the proximal tubules, which started to dilate at 48 hours after the obstruction and continued to develop over the first 7 days after the obstruction [15]. This study was designed to attempt to determine when this dilatation becomes irreversible. After the creation of a lower urinary tract obstruction, cystic dilatation of the ampullae in the nephrogenic zone
Fig. 7 In the bladder of a shunted lamb (Masson’s trichrome stain) (A and B), there is marked thickening of the bladder wall with fibrosis extending from the submucosal layer into the intramuscular region. On immunohistochemistry staining for SMA (C), myofibroblastic cell proliferation is recognized in the submucosal layer. The muscular layer is also thickened, but the muscular bands are fragmented by thick bands of collagen.
Vesicoamniotic shunt for urinary tract obstruction suppresses nephrogenesis [13]. This hypothesis may explain the relation between the timing of the obstruction and its effect on the developing kidney. Our previous study demonstrated that when we created an obstructive uropathy at 60 days, the nephrogenic zone is still intact 3 weeks after the obstruction [13]. If the nephrogenic zone is still intact, then our hypothesis was that the kidney would have a chance to produce new nephrons if the obstruction was relieved. In our model of obstructive uropathy created at 60 days in fetal lambs, we showed that in a series of nonshunted obstructed term lambs, half of the lambs had small kidneys with small numbers of dilated tubules in the cortex. However, 25% of the lambs in our previous study had bmacrocysticQ kidneys with prominent cystic dilatation of tubules in the mid to deep cortex [13]. Creating identical obstructions at identical times in renal development produced markedly different phenotypes of renal dysplasia [16]. We hypothesized that if a V-A shunt was able to release the pressure in the proximal tubules, the cystic dilatation will resolve, and new nephrons will be able to develop from the nephrogenic zone. We speculated that in the bmicrocysticQ kidneys, the cysts may also disappear, and new nephrons will continue to be produced until term. In human kidneys, the rate of development of glomerular populations during fetal life means that the total numbers of glomeruli rapidly increase before 24 weeks of gestation, and from 24 to 40 weeks, few new glomeruli are produced [17,18]. Our oligonephronia model had only small numbers of nephrons. This is related to this growth curve in the glomerular population. In the bmacrocysticQ kidneys in the shunted lambs, the cysts persist, and the dilated proximal tubules developed a cuff of connective tissue. Why this occurred is not clear but these 2 different types of kidney are somehow related to the dysplastic kidneys we reported in our previous paper in a nonshunted model [13]. Our results in this study suggest that a shunt operation for obstructive uropathy 3 weeks after an obstruction is produced in fetal lambs at 60 days of gestation is partially effective. The human equivalent of the timing of our shunting procedures is approximately 15 to 17 weeks of gestation. Most human shunting procedures are carried out at 20 to 24 weeks of gestation [19]. In human fetuses, the shunt-related complication rate is high if the shunts are performed at an early stage in gestation [8,20]. However, if optimal renal and bladder function are to be preserved, we are convinced that it is vital to ensure the development of normal bladder physiology by maintaining the normal voiding cycle. If an infant survives after the successful in utero placement of a V-A shunt, the next step is to preserve bladder function after shunting. This may be more difficult to assess than the assessment of the preservation of renal function. We reported the effect of early vesicostomy in obstructive uropathy on bladder development and found that after early shunting, the bladder wall was thickened with increased fibrosis in the submucosal layer [21]. These bladder abnormalities are probably the result of the initial
401 complete obstruction causing secondary changes to the development of the muscle walls in the bladder, resulting in an increased deposition of connective tissue [21]. Bladder hypertrophy is a common finding among patients with posterior urethral valves [22,23]. The hypertrophy is so marked that bladder compliance is significantly reduced. Our animal model demonstrated that the V-A shunted bladder can contain only 8.8 F 4.7 mL, and this is approximately one tenth of the capacity of the bladder in control animals. The bladder pressure-volume curve demonstrated that the V-A shunted model had extremely poor bladder compliance. In our nonshuntedterm lambs with obstructive uropathy, the severely dilated bladder demonstrated increased connective tissue deposition between the detrusor muscle bundles. After a V-A shunt for obstructive uropathy, the fetuses urinate through the shunt, and this procedure removes the normal bladder physiological cycle of filling and emptying against a slight resistance. The bladder capacity is very small because the shunt continuously drained the urine into the amniotic fluid. The problem is not only the size of the bladder but also the severely reduced bladder compliance. We used 2 different types of bypass procedure in this series of experiments. The vesicostomy was a surrogate for a V-A shunt but avoided the possibility of the shunt being dislodged. The placement of the perineal urethrostomy in the male lambs was an attempt to place a catheter that would descend below the external sphincter as the lamb grew, allowing the bladder to return to a normal filling and emptying cycle. Unfortunately, in all of the lambs, the catheter placed was too long and never descended below the external sphincter. This meant that it functioned in exactly the same way as the vesicostomy. In long-term follow-up of 14 children, Freedman et al [24] reported that 8 children void spontaneously, 4 children catheterise in addition to voiding, and 2 children catheterised only. Of the 4 children with posterior urethral valves, 3 have undergone bladder augmentation. Holmes et al [11] reported that 5 of the 8 survivors had some type of postnatal urinary diversion or augmentation. Our fetal lamb model is very similar to the posterior urethral valve patients, and this group might need bladder augmentation a few years after birth to increase bladder volume. From the histologic point of view, the SMA is transiently expressed in the myocardium and skeletal muscle during the development of the embryo and is highly restricted in adult animals to smooth muscle cells [25,26]. However, it is also expressed in myofibroblasts during wound healing and in a very limited number of adult tissues [27]. Little is known regarding the molecular mechanisms that control the activation of SMA in myofibroblasts in granulation tissue. On Masson’s trichrome stain, massive collagen deposition is observed in the bladder wall. Our immunohistochemical results demonstrated that the submucosal area represents proliferation of SMA-positive fibroblastic cells (bmyofibroblastsQ) in the walls of the shunted bladders, and
402 this results in the observed fibrosis in the bladder wall. Myofibroblasts are specialized contractile fibroblasts that are critical in wound closure and tissue contracture. From our animal data, the muscle hypertrophy and submucosal fibrosis are the key factors that result in the striking reduction in bladder compliance we have observed. Once severely dilated, the fetal bladder had increased connective tissue in the submucosal layer, extending into the intramuscular region. Muscle thickness was markedly increased in obstructed fetal bladders as the lambs approached term. If the muscle thickness increases with fetal age, an early shunt may preserve normal bladder compliance. Early shunting or the use of a pressure controlled shunt tube is the next phase in our studies.
Acknowledgment This study was supported by the Japanese Society for Grant aid for Scientific Research (C). Suture materials (Biosyn) were supplied by Tyco, Healthcare, Tokyo, Japan. These experiments would not have been possible without the skill and dedication of Doug Jensen of the Wallaceville Animal Research Facility. He consistently produces animals of timed gestation with the time of mating recorded to within 24 hours.
References [1] Golbus MS, Harrison MR, Filly RA, et al. In utero treatment of urinary tract obstruction. Am J Obstet Gynecol 1982;142:383 - 8. [2] Harrison MR, Golbus MS, Filly RA, et al. Fetal surgery for congenital hydronephrosis. N Engl J Med 1982;306:591 - 3. [3] Tanagho EA. Surgically induced partial urinary obstruction in the fetal lamb III. Ureteral obstruction. Invest Urol 1972;10:35 - 52. [4] Harrison MR, Nakayama DK, Noall R, et al. Correction of congenital hydronephrosis in utero. II. Decompression reverses the effect of obstruction on the fetal lung and urinary tract. J Pediatr Surg 1982; 17:965 - 74. [5] Glick PL, Harrison MR, Adzic NS, et al. Correction of congenital hydronephrosis in utero. IV. In utero decompression prevents renal dysplasia. J Pediatr Surg 1984;19:649 - 57. [6] Kitagawa H, Pringle KC, Zuccollo J, et al. The pathogenesis of dysplastic kidney in a urinary tract obstruction in the female fetal lamb. J Pediatr Surg 1999;34:1678 - 83. [7] Freedman AL, Johnson MP, Smith CA, et al. Long-term outcome in children after antenatal intervention for obstructive uropathy. Lancet 1999;354:374 - 7.
H. Kitagawa et al. [8] Freedman AL, Johnson MP, Gonzalez R. Fetal therapy for obstructive uropathy: past, present, . . .future? Pediatr Nephrol 2000;14:167 - 76. [9] Coplen DE. Prenatal intervention for hydronephrosis. J Urol 1997; 157:2270 - 7. [10] Freedman AL, Bukowski TP, Smith CA, et al. Fetal therapy for obstructive uropathy: diagnosis specific outcome. J Urol 1996;156:720 - 3. [11] Holmes N, Harrison MR, Baskin LS. Fetal surgery for posterior urethral valves: long-term postnatal outcome. Pediatrics 2001;108:E7. [12] Harrison MR, Filly RA. The fetus with obstructive uropathy: pathophysiology, natural history, selection, and treatment. In: Harrison MR, Golbus MS, Filly RA editors. The unborn patient: prenatal diagnosis and treatment. Philadelphia7 WB Saunders; 1991. p. 328 - 93. [13] Kitagawa H, Pringle KC, Koike J, et al. Optimal timing of perinatal treatment of obstructive uropathy in the fetal lamb. J Pediatr Surg 2003;38:1785 - 9. [14] Pringle KC. Fetal lamb and fetal lamb lung growth following creation and repair of a diaphragmatic hernia. In: Nathanielsz PW, editor. Animal models in fetal medicine. Ithaca (NY)7 Perinatology Press;1984. p. 109 - 48 [chapter 6]. [15] Kitagawa H, Pringle KC, Koike J, et al. The early effects of urinary tract obstruction on glomerulogenesis. J Pediatr Surg 2004;39:1845 - 8. [16] Kitagawa H, Pringle KC, Koike J, et al. Different phenotypes of dysplastic kidney in obstructive uropathy in fetal lambs. J Pediatr Surg 2001;36:1698 - 703. [17] Osathanondh V, Potter EL. Pathogenesis of polycystic kidneys. Survey of results of microdissection. Arch Pathol 1964;77:510 - 2. [18] Potter EL. Normal and abnormal development of the kidney. Chicago7 Year Book Medical Publishers, Inc; 1972. p. 15 - 21. [19] Clark TJ, Martin WL, Divakaran TG, et al. Prenatal bladder drainage in the management of fetal lower urinary tract obstruction: a systematic review and meta-analysis. Am J Obstet Gynecol 2003;102:367 - 82. [20] Jouannic JM, Hyett JA, Pandya PP, et al. Perinatal outcome in fetuses with megacystis in the first half of pregnancy. Prenat Diagn 2003; 23:340 - 4. [21] Sato Y, Kitagawa H, Pringle KC, et al. Effect of early vesicostomy in obstructive uropathy on bladder development. J Pediatr Surg 2004; 39:1849 - 52. [22] Bauer SB, Dieppa RA, Labib KK, et al. The bladder in boys with posterior urethral valve: a urodynamic assessment. J Urol 1979;121: 769 - 73. [23] Peters CA, Bolkier M, Bauer SB, et al. The urodynamic consequences of posterior urethral valves. J Urol 1990;144:122 - 6. [24] Freedman AL, Johnson MP, Smith CA, et al. Long-term outcome in children following antenatal intervention for obstructive uropathies. Pediatrics 1997;100:578. [25] Tomasek JJ, McRae J, Owens GK, et al. Regulation of alpha-smooth muscle actin expression in granulation tissue myofibroblasts is dependent on the intronic CarG element and the transforming growth factor-beta1 control element. Am J Pathol 2005;166:1343 - 51. [26] Hinz B, Gabbiani G. Mechanisms of force generation and transmission by myofibroblasts. Curr Opin Biotechnol 2003;14:538 - 46. [27] Darby I, Skalli O, Gabbiani G. Alpha-smooth muscle actin is transiently expressed by myofibroblasts during experimental wound healing. Lab Invest 1990;63:21 - 9.