Bladder autoaugmentation in the rabbit using de-epithelialized segments of small intestine, stomach and lyophilized human dura mater


Summary

Objective

To develop an animal model of partial detrusorectomy (autoaugmentation) and thus avoid the consequences of the direct contact of intestinal mucosa with the urinary tract in bladder augmentation.

Materials and methods

A diverticular urothelial bulge was created and patched with demucosalized segments of small bowel (group A), stomach (group B) and with lyophilized human dura mater (group C). The surgery was performed on 50 New Zealand rabbits which were compared with 10 control animals and killed at 2, 4 and 6 weeks after surgery. Urodynamic studies and cystography were performed before operation and at death, and the augmented bladders examined histologically.

Results

Six weeks after the procedure, the mean (SD) bladder compliance was 22.7 (5.7) in group A (intestinal patch, n = 6), 2.3 (0.5) in group B (stomach patch, n = 3), 3.1 (1.9) in group C (lyophilized human dura, n = 3) and 9.4 (0.4) in the control group (n = 4). Histological studies showed residual enteric and gastric mucosa but an intact urothelium under the intestinal patch.

Conclusion

The results of this experimental study suggest that a demucosalized segment of small bowel is the best material to increase bladder compliance in detrusorectomy (autoaugmentation) as applied in this animal model.

Introduction

The purpose of bladder augmentation is to increase the storage capacity and develop low storage pressure. The indications for augmentation include a non-compliant bladder caused by tuberculosis or interstitial cystitis, neurogenic bladder and bladder carcinoma [1][2]. To date, enterocystoplasty has been effective in augmenting the bladder, but complications can arise from the immediate contact of urine with the intestinal tract [3]. The most severe complications are electrolyte disturbances and shifts [4]. Other serious complications are mucus production, stone formation, enteric fistulae, abscess, peritoneal adhesions and the potential for malignancy [5][6].
Different gastrointestinal segments have also been used, including stomach and colon, which can cause serious complications and increased post-operative morbidity [7]. As an alternative method of augmenting the bladder, synthetic materials have been used experimentally as patches on the incised bladder, but the results were not always satisfactory [8]. Because there are complications associated with these methods, detrusorectomy (autoaugmentation) was developed, the basic concept of which is to create a large bladder diverticulum by removing the detrusor muscle almost entirely over the dome of the bladder, and to leave the underlying epithelium intact [9][10]. The clinical results of this method have not been satisfactory in the long-term because fibrosis develops in the diverticulum [11][12]. Demucosalized patches of colon and stomach have been used by other investigators in different animal models to cover the surface of the diverticulum and avoid fibrosis [12].
In the present experiment in a rabbit model, isolated segments of demucosalized small bowel and stomach were used as patches attached to the bladder diverticulum and sutured onto the borders of the dissected detrusor muscle. The large bulging epithelial surface was thus secured and the underlying epithelium allowed to remain intact, avoiding the undesirable consequences of incorporating gastrointestinal mucosa into the urinary tract. Apart from intestine and stomach, lyophilized human dura mater (LHD) was used as an alternative natural material to cover the diverticulum [13][14]. We describe the results and surgical technique applied in a rabbit model using this method of partial detrusorectomy.

Materials and methods

The study included 50 young male New Zealand rabbits (body weight 1.5-3.5 kg). A cystogram was performed pre-operatively on each rabbit, under anaesthesia using a mixture of 10 mg/kg each of ketamine and promazine-HCL injected subcutaneously. Urographin (58%) was injected into the bladder through an 8 F catheter and radiographs obtained in posteroanterior and lateral views. The same anaesthetics were used for the surgical procedures, supplemented intra-operatively as necessary. Cefamandole nafate 50 mg/kg and metronidazole 25 mg were administered intraperitoneally both pre-operatively and immediately before the completion of the operation; no antibiotics were given post-operatively. Intravenous lines were established pre-operatively and 5% dextroselactate Ringer’s solution administered at 30 mL/h.
The abdomen was shaved, prepared and the rabbits placed supine on the operating table. The peritoneum was entered through a midline incision, the bladder exposed and punctured with an 18 G needle. An 8 F catheter (paediatric feeding tube) was passed through the urethra and the bladder was emptied. It was then filled through the 18 G needle with normal saline at 20 mL/min until the saline leaked at the external meatus. The intravesical pressure was measured at the leak point using a central venous pressure system and the volume measured by emptying the bladder completely. The puncture site was closed with 4/0 chromic catgut. Using x4 magnification, an incision of the detrusor was gently spread laterally with blunt dissection until a diverticulum of the underlying bladder epithelium was exposed along the entire incision, to about one-third the area of the bladder. The dissected detrusor flaps on each side were excised. The animals were divided into three groups; the bladder diverticulum was covered in group A with a de-epithelialized segment of small intestine, in group B with a de-epithelialized segment of stomach and in group C with a patch of LHD.
In group A, a 3-5 cm (according to the size of the diverticulum) segment of the small intestine was isolated with an intact vascular pedicle. The continuity of the bowel was re-established using end-to-end anastomosis in two layers with running 4/0 chromic cat-gut sutures for the anastomosis and interrupted sutures for the seromuscular layer. The isolated bowel was then irrigated with 0.5 x normal saline and was opened along its antimesenteric border, creating a rectangular segment and exposing the mucosal surface. Using a scalpel blade, the mucosa was removed by scraping it from the centre to the periphery of the bowel segment. Haemostasis was achieved by direct pressure or electrodiathermy when necessary. The bladder was then distended by filling to near capacity with an 8 F catheter, the bowel segment sutured to the excised lateral edges of the detrusor and its posterior and anterior corners with 4/0 polyglactin. The isolated bowel was completely secured on the mucosal diverticulum with interrupted sutures along the remaining borders of the detrusor. With the bladder fully distended, any escape of the thin mucosal diverticulum between sutures was identified and extra sutures used to close the gaps between detrusor and intestinal patch (Fig. 1).
Covering of the bladder mucosa diverticulum
Fig. 1. Covering of the bladder mucosa diverticulum with an isolated intestinal segment.
In group B, after measuring the volume and intravesical pressure, the midline incision was extended to the upper abdominal wall and the stomach exposed. The bladder and intestine were covered with gauze pads soaked in normal saline at body temperature. An oval section of the greater curvature was clamped, including the anterior and posterior wall of the stomach, and a 4-6.5 cm segment excised with an electrocautery blade, leaving intact its blood supply from the left gastroepiploic vessels. The isolated segment was then irrigated with 0.5 x normal saline and haemostasis achieved mainly by direct pressure. The gastric mucosa was de-epithelialized by scraping with a scalpel. The stomach was closed in two layers with respective running and interrupted sutures of 4/0 polyglactin. The excised segment was well mobilized and passed through the mesentery to reach the bladder, and its serosal layer sutured to the borders of the excised detrusor with 4/0 polyglactin interrupted sutures.
In group C, the same procedure was followed to create the bladder mucosa diverticulum and with the bladder distended, a piece of LHD was cut to size and sutured to the detrusor edges with interrupted 4/0 polyglactin. The catheter was removed at the end of the procedure. No drains were left and the wound was closed in two layers.
All animals were allowed to recover and fed the usual laboratory diet except those in group B which were fed minimal maintenance portions for the first week post-operatively as a precaution against stomach perforation. Some animals of all groups were killed at 2, 4 and 6 weeks; before death, cystography was performed under anaesthesia. Through a midline incision, the bladder capacity and intravesical pressure were measured as described previously, and with the bladder filled with 10% formalin, cystectomy was performed. After the bladder was examined grossly, sections of the specimen were cut, stained with haematoxylin and eosin, and examined histologically.
The significance of differences in bladder compliance was tested using a two-way ANOVA with group (A, B, C or D) and time of death (2, 4 or 6 weeks) as factors. To account for possible effects caused by differences in pre-operative conditions, an analysis of covariance was also carried out with pre-operative compliance as the covariate. Both methods were also applied to the data after logarithmic transformation; the significance levels differed slightly but all results remained unchanged. The statistical methods used are detailed elsewhere [15].

Results

Of the 22 animals in groups A, 10 died from enteric anastomotic haematoma during the first post-operative day; of the 18 in group B, two died intra-operatively and another seven animals died from stomach perforation 5 days after surgery; of the 10 with the bladder diverticulum covered by LHD, one animal died from unknown causes. Of a total of 50 animals operated, 30 survived to be evaluated urodynamically and histologically, and for comparison with the 10 control animals.
Cystograms taken 6 weeks after surgery
Fig. 2. Cystograms taken 6 weeks after surgery showing, a, marked enlargement of the bladder in group A (intestinal patch), b, decrease in the bladder in group B (gastric segment patch) and, c, intense shrinkage and obliteration of the dome of the bladder in group C (LHD patch).
The pre-operative retrograde cystograms showed bilateral VUR in many animals and a smooth bladder contour. During the sixth week, there was a marked enlargement of the bladder in the animals in group A (Fig. 2a). Post-operative cystograms of the animals in groups B and C showed a decrease in the bladder at 6 weeks, with intense shrinkage (Fig. 2b) and obliteration of the contour of the dome of the bladder, respectively (Fig. 2c).
Table 1 The mean bladder compliance by group and sample time
Mean (SEM) bladder compliance (mL/cmH2O) [n]
Sample (weeks) Group A Group B Group C Group D
2 5.2 (3.5) [3] 8.5 (3.6) [3] 4.1 (0.6) [3] 11.2 (1.6) [3]
4* 27.4 (10.7) [3] 4.0 (2.8) [3] 3.7 (0.6) [3] 12.2 (0.4) [3]
6 22.7 (5.7) [6] 2.3 (0.5) [3] 3.1 (1.9) [3] 9.4 (0.4) [4]
*P<0.01 for the group interaction in a two-way ANOVA, not significant for sample time and interaction
The mean (SEM) bladder compliance for the four groups at the three sample times is shown in Table 1; the differences among the groups were significant (F = 6.907; d.f. 3,28; P<0.01), but those among the sample times were not (F = 0.954; d.f. 2, 28; P = 0.398). Group A had significantly higher mean compliances than the other groups, groups B and C were similar and low, while group D were intermediate. The interaction effect was marked but not significant (F = 2.036; d.f. 6,28; P = 0.094) and was related to the slightly different group differences at 2 weeks. The analysis of covariance yielded similar results, even though the covariate had a significant positive effect (F = 7.699; d.f. 1,27; P<0.01). These differences were also confirmed by the one-way ANOVA with group as the only factor (F = 7.204; d.f. 3,39; P< 0.001). The multiple comparison procedure of Neuman-Keuls indicated that groups B and C had similar means, significantly smaller than those of both group A and D; group A had a significantly larger mean value than group D.
Sections at 6 weeks from different pathces
Fig. 3. Sections at 6 weeks from a, the intestinal patch (group A), showing well-attached urothelium covering the residual intestinal mucosa and no evidence of detrusor fibres, b, the stomach patch (group B), showing fragmented urothelium covering the residual gastric mucosa and no detrusor fibres; and, c, the LHD patch, showing preserved urothelium with profound inflammatory changes and no residual bladder musculature. All sections, haematoxylin and eosin. x 200.
Histologically, there was evidence of intestinal mucosa with sparse foci of denudation in the animals in group A. A urothelium 4-6 cells thick covered all areas of the patch and the intestinal layer appeared to be well attached to the bladder mucosa. There was mild inflammatory subepithelial infiltrate at the edges of the patch, especially near the suture sites. No detrusor fibres were found in any section of the patch (Fig. 3a).
Gastric mucosa was also present in the animals in group B, with transitional epithelium overlying the gastric mucosa along its luminal side. The urothelium was fragmented at different sites and the urothelial lining of the bladder contracted, with residual inflammation noted near the edges of the patch. There was no evidence of bladder muscle fibres in any section (Fig. 3b).
In almost all the animals in group C, there was profound chronic inflammation between the patch and the urothelium, with detachment of the LHD layer from the oedematous subepithelial tissue, especially at 4 and 6 weeks. Although the urothelium was preserved along the luminal side of the patch, there was intense calcification and stone formation, with marked fibrosis and contraction of the urothelium near the outward boundaries of the patch. No residual bladder musculature was found in any of the sections (Fig. 3c).

Discussion

Augmentation using segments of bowel was first described almost a century ago by Mikulicz in 1899 [16]. Various segments of intestine and stomach have been used since then to augment the bladder. Enterocystoplasty has gained wide acceptance but may cause electrolyte disturbances arising from the direct contact of urine with the intestinal mucosa.
Various investigators have previously used demucosalized segments of intestine for enterocystoplasty in rats, dogs and calves in an attempt to cause regrowth of the urothelium over the surface of the demucosalized muscle, with encouraging results in rats but poor results in dogs and calves [9][10][17][18][19][20][21]. The concept of autoaugmentation is simple; a large diverticular bulge is created by stripping the detrusor, with dissection at the dome of the bladder, increasing the storage capacity of the bladder and decreasing storage pressure if the diverticulum maintains its elastic properties in the long-term. It would appear that for the urothelial diverticulum to persist, a muscular backing should be provided that preserve the augmenting effect. Using demucosalized segments of stomach as a patch for autoaugmentation gastrocystoplasty in sheep and humans resulted in a satisfactory augmenting effect [12][22].
The methods in the present study avoided the consequences of incorporating gastrointestinal mucosa into the urinary tract and used readily available urothelium as an inner lining of the gastrointestinal segments. It was very difficult to remove the mucosa and submucosa to avoid regeneration of the intestinal epithelium [3][17][23]. The small bowel of the rabbit is thin and friable and cannot tolerate demucosalization [22][24]. Although normal saline was injected into submucosa, it was virtually impossible to find a plane of cleavage between submucosa and muscle so that the submucosal-mucosal layer could be peeled away. When this was tried the intestinal patch was torn, causing haemorrhage and haematoma. The mucosa was de-epithelialized by scraping with a scalped blade and bleeding was easily controlled by applying direct pressure. Removing the mucosa from the stomach was also difficult, although it seems that stomach is an ideal material for demucosalization with diathermy dissection in augmentation gastrocystoplasty in humans [12].
In the present experimental model, finding a suitable size of stomach segment to fit the bladder diverticulum was always difficult, because in the rabbit the proportion of urinary bladder to stomach is at least 2 : 1, a value further increased with the creation of the diverticular bulge. The integrity of the patch had to be preserved [2] and therefore demucosalization with diathermy dissection could not be used because it would cause fibrosis and contracture by thermal damage [11][17][18]. Another reason for avoiding this denuding technique was that it prolongs the operating time and increases post-operative morbidity [24][25]. Thus, the mucosa was de-epithelialized by scraping with a scalpel and avoiding excessive haemorrhage.
Although the histological examination showed evidence of residual intestinal mucosa, the urodynamic results showed a significant improvement in bladder capacity and compliance 6 weeks after surgery in group A. Macroscopically, the diverticulum seemed to maintain its elasticity and smooth contour, and microscopically the bladder diverticular urothelium was completely adherent to the intestinal muscle layer. However, this was not so for the animals in group B; at 2 and 6 weeks there was marked shrinkage of the patch and urodynamics showed reduced bladder capacity and compliance. Although gastric muscle used in sheep and humans has shown good results [12][23][26], in the present experimental model the effect was poor probably because of the increased contractility of the segments applied. The rabbits in group C had even worse fibrosis and contracture after 2 weeks, and stone formation at 6 weeks. These findings suggest that LHD is completely unsuitable for bladder autoaugmentation.
In conclusion, in this rabbit model the best material for augmentation was a segment of small bowel; although it cannot be completely demucosalized, the epithelium of the bladder diverticulum was maintained intact and retained its elastic properties, producing a good augmenting effect. The segments of stomach and LHD caused fibrosis and graft contracture, and are unsuitable materials for this technique in this rabbit model. Perhaps in future less traumatic procedures will be developed for removing the mucosa which might produce better results. The successful outcome with the use of small bowel segments justifies further experimental research which may lead to future clinical application.

Acknowledgements

We are grateful to Professor G. Eliopoulos, Assoc. Prof. I. Vlachonikolis and M. Mavromanolakis (Biostatistics Lab) for their valuable assistance in statistical analysis. We also thank Dr K. Tsagaraki for her excellent technical assistance in the X-ray studies.

Authors

  • A. Cranidis, MD, PhD, Professor and Chairman.
  • G. Nestoridis, MD, Consultant Urologist.
  • D. Delakas, MD, Assistant Professor.
  • P. Lumbakis, MD, Senior Registrar.
  • P. Kanavaros, MD, Assistant Professor.
  • Correspondence: Dr A. Cranidis, Department of Urology, University General Hospital, PO Box 1352, Heraklion 71110, Crete, Greece.

References

  1. ^ Hendren WH, Hendren RB. Bladder augmentation: experience with 129 children and young adults. J Urol 1990; 144: 445-3
  2. ^a ^b Shoemaker WC, Marucci HC. The experimental use of seromuscular grafts in bladder reconstruction: preliminary report. J Urol 1955; 73: 314-21
  3. ^a ^b Blandy JP. The feasibility of preparing an ideal substitute for the urinary bladder. Ann R Coll Surg Engl 1964; 35: 287-311
  4. ^ Koch MO, Mc Dougal WS. The pathophysiology of hyperchloremic metabolic acidosis after urinary diversion through intestinal segments. Surgery 1985; 98: 561-70
  5. ^ Filmer BR, Spencer JR. Malignancies in bladder augmentations and intestinal conduits. J Urol 1990; 143: 671-8
  6. ^ Smith P, Hardy GT. Carcinoma occurring as a late complication of ileocystoplasty. Br J Urol 1971; 43: 576-9
  7. ^ Muraishi O, Ikado S, Yamashita I, Yamaguchi K, Ogawa A. Gastrocystoplasty in dogs an ulcerating effect of acid urine. J Urol 1992; 147: 242-5
  8. ^ Kelami A, Dustmann HO, Lodtke-Handjery A, Carcamo V, Herold G. Experimental investigations of bladder regeneration using Teflon-felt as a bladder wall substitute. J Urol 1970; 104: 693-8
  9. ^a ^b Cartwright PC, Snow BW. Bladder auto-augmentation: partial detrusor excision to augment the bladder without use of bowel. J Urol 1989; 142: 1050-3
  10. ^a ^b Cartwright PC, Show BW. Bladder auto-augmentation early clinical experience. J Urol 1989; 142: 505-8
  11. ^a ^b Snow BW, Cartwright MD. Bladder augmentation. In King LR, Stone AR, Webster GD eds, Bladder Reconstruction and Continent Urinary Diversion. Chicago: Year Book Medical Publishers, 1991: 107-14
  12. ^a ^b ^c ^d ^e Dewan PA, Stefanek W. Autoaugmentation gastrocystoplasty: early clinical results. Br J Urol 1994; 74: 460-4
  13. ^ Kelami A. Lyophilized human dura as a bladder wall substitute: experimental and clinical results. J Urol 1971; 105: 518-22
  14. ^ Novick AC, Stratton RA, Banowsky LH, Nose Y, Levin H, Stewart BH. Experimental bladder substitution using biodegradable graft of natural tissue. Urology 1977; 10: 118-27
  15. ^ Armitage P, Berry G. Statistical Methods in Medical Research. 2nd Edn, Oxford: Blackwell, 1989
  16. ^ Mikulicz L. Zur Operation der Augeborenen. Belsasens palte Zentralbl F Chir 1899; 26: 641-8
  17. ^a ^b ^c Motley RC, Montgomery BT, Zollman PE, Holley KE, Kramer ST. Augmentation cystoplasty utilizing de-epithelialized sigmoid colon: a preliminary study. J Urol 1990; 143: 1257-60
  18. ^a ^b Salle TLP, Fraga TCS, Lucib A, Lampertz M, Jobim G, Putten A. Seromuscular enterocystoplasty in dogs. J Urol 1990; 144: 454-6
  19. ^ Oesch I. Neurothelium in bladder augmentation. Eur Urol 1988; 14: 328-9
  20. ^ Badiola de F, Manivel JC, Gongales R. Seromuscular enterocystoplasty in rats. J Urol 1991; 146: 559-62
  21. ^ Blandy JP, Mc Donald JH. Heterotopic ossification in uroepithelial regeneration on grafts of ileum and colon. Surg Forum Urol 1961; 12: 498-500
  22. ^a ^b Dewan PA, Stefanek W. Auto-augmentation colocystoplasty. Pediatr Surg Int 1994; 9: 526-8
  23. ^a ^b Dewan PA, Byard RW. Auto-augmentation gastrocystoplasty in a sheep model. Br J Urol 1993; 72: 56-9
  24. ^a ^b Dewan PA, Lorenz C, Stefanek W, Byard RW. Urothelial lined colocystoplasty in a sheep model. Eur Urol 1994; 26: 240-6
  25. ^ Njnyen HD, Mitchell EM, Horowitz M, Bagli JD, Carr CM. Demucosalized augmentation gastrocystoplasty with bladder autoaugmentation in pediatric patients. J Urol 1996; 156: 206-9
  26. ^ Snow BW, Cartwright MD. Bladder autoaugmentation. Urol Clin North Am 1996; 23: 323-31

Author

Sandy Raharja Sandy Raharja UrologyJournal.US is one of the most highly respected medical journals on the web. It presents unparalleled coverage of contemporary urological opinion, clinical reports and studies, innovations and developments. Every issue gives invaluable practical information in the form of primary papers, review articles, case reports, basic science articles, correspondence and book reviews and reports of meetings.

Post a Comment