Ordinary Glass Spheres as an Alternative Injectable Material for Endoscopic Correction of Vesicoureteral Reflux

0022-5347/04/1713-1282/0 THE JOURNAL OF UROLOGY® Copyright © 2004 by AMERICAN UROLOGICAL ASSOCIATION

Vol. 171, 1282–1286, March 2004 Printed in U.S.A.

DOI: 10.1097/01.ju.0000112791.63272.6c

ORDINARY GLASS SPHERES AS AN ALTERNATIVE INJECTABLE MATERIAL FOR ENDOSCOPIC CORRECTION OF VESICOURETERAL REFLUX FERRUH SIMSEK, SELCUK YUCEL,* MUSTAFA AKTAS

AND

LEVENT TURKERI

From the Department of Urology, Marmara University School of Medicine, Marmara and Department of Surgery, Istanbul University School of Veterenary Medicine, Istanbul, Turkey

ABSTRACT

Purpose: Although they are therapeutically effective, injectable materials for urinary tract are associated with various disadvantages, precluding their universal acceptance. In this study we investigated glass spheres (GSs) as an alternative injectable substance to correct vesicoureteral reflux (VUR) in an animal model. Materials and Methods: We used 150 to 300 ␮ GSs suspended in agarose gel to form the injection paste. GS paste was injected into the rectus muscle and submucosa of the bladder in 8 adult New Zealand male rabbits. As a control group, vehicle only was injected into 4 rabbits. The rabbits were sacrificed to harvest the bladder, pelvic lymph nodes, kidney, liver, brain, spleen and lung at month 1 and year 1 of injection. A VUR model was then created by unroofing the 2 ureteral orifices of 12 adult sheep. GS paste was injected into the right subureter and vehicle only was injected into the left subureter. Cystourethrographies were performed at month 3 and year 1 of injection. The sheep were sacrificed at cystourethrography to harvest the bladder, lymph nodes, kidney, liver, brain, spleen and lung. Results: At month 3 and year 1 of injection into rabbit tissues nodule formation was stable in position and volume. Histopathological studies of local and distant organs of the rabbit did not show any granuloma formation or migration of GS. GS paste injection corrected VUR in sheep. Re-injection of GS into still refluxing left units corrected VUR. Local and distant organs harvested from sheep did not demonstrate distant migration. Conclusions: When injected into bladder submucosa and rectus muscle, GS appears to be inert, biocompatible and efficient. Similarly it is effective for correcting VUR in an animal model. We present our data on GS, encouraging further investigation to develop an alternative injectable material for endoscopic VUR correction. KEY WORDS: urethra, vesico-ureteral reflux, glass, rabbits, sheep

Submucosal endoscopic injection for the correction of vesicoureteral reflux (VUR) is a safe, comfortable and low cost operation compared to open antireflux surgery.1 However, the major issue impeding its universal acceptance is absence of a convenient injection material. Ideal injectable material for the urinary tract has been defined as nonmigrating, durable, biocompatible, nontoxic, noncarcinogenic, nonteratogenic, easily injectable and affordable.2 We hypothesized about an inexpensive injectable material, namely glass spheres (GSs) and investigated its migration, safety and efficacy profile in animal experiments. Nonautologous substances, including polytetrafluoroethylene (PTFE),3 glutaraldehyde cross-linked bovine collagen,4 polydimethylsiloxane,5 detranomer in sodium hyaluronan,6 self-detachable balloon7 and autologous substances, such as cultured bladder muscle cells,8 fibrin,9 fat,10 collagen11 and chondrocytes,12 have been used. PTFE, which is the most famous injection material in the recent past,3 is largely out of use because of its migration potential, causing embolization or granuloma formation.13 Dextranomer in sodium hyaluronan has gained more popularity with remarkable tissue augmenting capacity and initial success in VUR correction but long-term durability and efficacy remain under investiga-

tion.6 Although the idea of using autologous material for injection treatment is appealing because of the low immunogenic reaction, the main drawback is rapid volume loss due to absorption by the host body.2 Unfortunately, as investigations deepen to discover the most ideal injectable material, substance affordability significantly decreases. For instance, while the first injection material in Turkey, PTFE, is $100, recently developed materials are more expensive, such as polydimethylsiloxane ($580), bovine dermal collagen in phosphate buffered saline ($1,500) and dextranomer in sodium hyaluronan ($900). All recently improved injection materials are exclusively expensive, hampering their widespread use in low income countries. In this study we prepared a cheap GS injection paste to investigate its local and distant tissue reactions after submucosal and intramuscular injections as well as its efficacy for the endoscopic correction of iatrogenic VUR in an animal model. MATERIALS AND METHODS

Injection material preparation. GSs 150 to 300 ␮ in diameter were used (Turkish Glass Works Co., Istanbul, Turkey). Figure 1 shows toluidine blue stained GSs under light microscopy. GSs were sterilized in an ethylene oxide gas autoclave. Injection paste was formed by mixing 2 gm GSs and sterile 10 cc 2% agarose gel. The sterile mixture was homogenized by heating the gel to 60C in a hot water bath. The

Accepted for publication October 24, 2003. Study received committee on animal research approval. Supported by Research Foundation of Marmara University. * Correspondence: Department of Urology, Akdeniz University School of Medicine, Kampus 07070, Antalya, Turkey (telephone: 90242-227-4480; FAX: 90-242-227-4482; e-mail: [email protected]). 1282

GLASS SPHERES AS INJECTABLE MATERIAL FOR VESICOURETERAL REFLUX

FIG. 1. Toluidine blue staining of GS paste shows regular, uniform appearance of GSs. Bar represents 150 ␮m.

homogenized mixture was left to cool at room temperature to obtain the viscous homogenous injection paste. Table 1 shows the chemical components of glass. Animal experiments. Rabbits: The distant migration potential and local tissue reaction of GS were investigated. We used 12 adult New Zealand male rabbits weighing 5 to 7 kg. Four rabbits served as the control group. They were anesthesized with 10 mg/kg ketamine hydrochloride and 0.02 mg/kg xylazine hydrochloride intramuscular injection. The bladder was opened to visualize the bladder mucosa. GS paste (0.5 cc) was introduced through a 23 gauge needle and 1 ml syringe into the submucosa of the posterior bladder wall to produce significant bulging. After bladder closure 0.5 cc injection paste were injected into the rectus muscle using the same needle. Control rabbits were treated in the same manner with the injection of agarose gel only into the rectus muscle and submucosa of the posterior bladder wall. A single dose of 25 mg/kg ampicillin/sulbactam was given intramuscularly as prophylactic antibiotic therapy. At month 1 after operation 4 rabbits in the GS injected group and 2 in the control group were sacrificed by CO2 asphyxia. The bladder, liver, brain, spleen, lung, kidney, rectus muscle and nearby lymph nodes were harvested. Careful macroscopic examination was performed of the bladder and rectus muscle injection area to assess the size of the bulging sites and local tissue reaction, which were photographed. Similarly other organ and lymph nodes were macroscopically examined for any sign of injected material migration and photographed. Organs were sliced 1 mm thick by a sharp brain knife and examined by a hand lens for the pathological appearance. All harvested tissue was fixed in 10% formalin solution and sectioned at 10 ␮ for hematoxylin and eosin, and Masson’s trichrome staining for histological examination. At least 3 sections per tissue were randomly selected for histopathological examination. At year 1 after injection the remaining rabbits were sacrificed and examined macroscopically and histologically, as described. Sheep: The efficacy of GS paste injection to correct VUR was investigated. We used 12 adult sheep weighing 40 to 60 kg as an iatrogenic VUR model. Sheep were anesthesized by intramuscular injection of 5 mg/kg ketamine hydrochloride

1283

and 0.02 mg/kg xylazine hydrochloride. Cystourethrography (CUG) revealed that all 12 sheep were VUR-free. After the bladder was emptied, a 24Fr offset angle lens endoscope was transurethrally introduced into the bladder. Following the unroofing of the 2 ureteral orifices at the 12 o’clock position with ureteral orifice scissors CUG was repeated. All 24 unroofed ureteral units showed grades II to III VUR. As described by O’Donnell and Puri,14 GS injection paste was introduced submucosally to the right ureter via a PTFE needle using a PTFE injection gun through the same 24Fr offset angle lens endoscope in 12 unroofed sheep. The amount of injected GS material was 0.8 to 1.2 cc. Agarose (1 cc) gel was injected in the same manner into the left subureter of the same sheep as a control. Prophylaxis was done as described. All unroofed and injected sheep survived. At month 3 CUG was repeated in 12 sheep under general anesthesia, of which 6 were sacrificed by intravenous 26% pentobarbital injection to harvest the bladder, pelvic lymph nodes, liver, lung, brain, spleen and kidney for macroscopic and microscopic examination. In the bladder the swelling of GS injection was specifically assessed and photographed. Tissues were fixed, sectioned, stained and evaluated as described. In 3 of the remaining 6 sheep reinjection of GS paste was performed into the still refluxing left ureter, as described. At year 1 after injection CUG was done again. The 6 sheep were sacrificed, harvested, and examined macroscopically and microscopically, as described. RESULTS

GS particles in injection paste. As stated by the manufacturer, 150 to 300 ␮m GSs were almost uniform in the paste (fig. 1). The injection paste was so viscous that it could be easily injected by the PTFE injection gun through an 18 gauge PTFE needle. Local tissue reaction. Rabbits: In the 4 GS injected rabbits at month 1 the reaction to GS paste at the injection site (rectus muscle and submucosa of the bladder) was minimal. A thin fibrous capsule had formed around the GS aggregates in the rectus muscle and submucosa (fig. 2, A). No GS particles were found outside of the fibrous capsulated focus. Macroscopically the bulging of injected GS paste in the rectus muscle and bladder submucosa at month 1 was stable compared to its appearance at the first injection session (fig. 3,

TABLE 1. Chemical composition of GSs Element

% Mass

SiO2 Na2O CaO MgO Al2O3 K2O SO3 Fe2O3 TiO3

70.64 14.07 9.64 3.96 0.94 0.28 0.28 0.18 0.037

FIG. 2. Fibrous capsule (c) around GS (gs) paste injection in submucosa of rabbit bladder. A, early formation of fibrous capsule without any distinctive giant cell formation at month 1 after injection. B, late appearance of fibrous capsule without any distinctive giant cell migration and granuloma formation at year 1 after injection. Bar represents 150 ␮m.

1284

GLASS SPHERES AS INJECTABLE MATERIAL FOR VESICOURETERAL REFLUX

A). The control group in which agarose gel alone was injected did not show any histological derangement. However, in the 4 GS injected rabbits at year 1 after injection GSs were surrounded by a thicker fibrous capsule without any significant surrounding inflammatory cells (fig. 2, B). No giant cell formation was noted at year 1. The volume of the GS injection bulge was the same in the submucosa and rectus muscle at year 1 (fig. 3, B). In the control group the histological appearance was normal. Sheep: In the 9 ureteral units GS injected the reaction to GS paste at the injection site was minimal at month 3. A thin fibrous capsule had formed around the GS cluster in the bladder submucosa. Macroscopically bulging of the injected GS paste in the bladder submucosa at month 3 was stable compared to its appearance at the first injection session. The 6 left ureteral units in which agarose gel alone was injected did not show any different histological picture compared to normal units. However, in the sheep with GS injected into the 6 right ureteral GSs were surrounded by a fibrous capsule composed of collagenous fibers without any pronounced surrounding chronic inflammatory cells at year 1 (fig. 4). No giant cell formation was noted at year 1. Macroscopically the bulging volume was similar at year 1. In the control group, similar to the month 3 examination, histological evaluation did not reveal any histopathological change. Distant sites and migration potential. Rabbits: At month 1 GSs were not noted at distant sites (adjacent lymph nodes, kidney, brain, lung, spleen and liver) in rabbits with GS paste injected into the rectus muscle and bladder submucosa. Likewise at year 1 of injection no histological sign of migration of injected GS to the same harvested organ was found. Sheep: At month 3 no migrated GS particles were observed in harvested organs (pelvic lymph nodes, kidney, brain, lung, spleen and liver) in all sheep with GS paste injected into the bladder submucosa. At year 1 after injection no histological sign of distant migration of injected GS was noted. Efficacy of GS paste injection for vesicoureteral reflux. Table 2 shows that unroofing was found an effective method for achieving iatrogenic VUR. At month 3 after GS paste injection in 10 of the 12 right ureteral units VUR was observed to be resolved, whereas the 12 left ureteral control units with agarose gel injected were still refluxing. The second injection

FIG. 4. Collagenous capsule (c) between injected GS paste and bladder mucosa (m) of sheep at year 1 after injection. Bar represents 100 ␮m. TABLE 2. VUR in sheep after unroofing and injections No. Ureteral Units (%)

Unroofing: Before After After GS injection: 3 Mos 1 Yr After vehicle injection After vehicle ⫹ GS injection

Refluxing

Nonrefluxing

Totals

0 24 (100)

24 (100) 0

24 (100) 24 (100)

2 (16.6) 0 12 (100) 0

10 (93.4) 6 (100) 0 3 (100)

12 (100) 6 (100) 12 (100) 3 (100)

of GS paste into 3 of the 6 left ureteral units still refluxing completely resolved the reflux. However, the remaining 3 control units were still refluxing at year 1. The nonrefluxing right ureters into which GS paste had been injected were still nonrefluxing at year 1 (table 2, fig. 5). DISCUSSION

FIG. 3. Bulging (B) of GS paste after injection into rabbit bladder submucosa (b). A, bulging is significant on bladder wall at month 1 after injection. B, bulging is still in same shape without any volume loss at year 1 after injection. Bar represents 1 mm.

Pediatric urologists greatly appreciated the success of the endoscopic approach for VUR.1 However, to our knowledge since no ideal injection material has been discovered, serious concerns about the safety and efficacy of injectable materials ensued. The ideal substance for subureteral injection should be inert, nonmigrating, noncarcinogenic, nonteratogenic, nonimmunogenic and inexpensive.2 Injected material should produce a durable mass that retains its volume and position to provide a solid support to anchor and compress the intravesical ureter.2 In this study we investigated the feasibility of a hypothesized GS paste as an inexpensive, inert and durable material for injection therapy. Nonautologous injection substances are the most common materials since their absorption potential is slight compared to autologous ones. However, the main issues with nonautologous materials are remarkable foreign body reaction and migration risk. For example, the most popular injection material, PTFE, is not preferred by most pediatric urologists because of granuloma formation and distant migration potential.13 Most developed materials reveal a significant local tissue reaction, characterized by giant cells after injection.13, 15–18 Industrially manufactured substances are supported with noninert chemicals to avoid displacement by local tissue integration. Glass is a naturally occurring, inexpensive and inert material with a minimal foreign body re-

GLASS SPHERES AS INJECTABLE MATERIAL FOR VESICOURETERAL REFLUX

FIG. 5. Voiding CUG in sheep. A, before unroofing to create VUR. B, after unroofing to create VUR bilateral reflux. C, at month 3 after GS injection into right subureter only left ureter refluxed. D, at year 1 after first injection 2 ureters were nonrefluxing following GS injection into left subureter.

action. In our experiment we did not note any giant cell formation at the injection site at short-term and long-term followup of the rabbits (figs. 2 and 4). Ordinary glass is composed of 12% sodium oxide (Na2O), 12% calcium oxide (CaO) and 76% silicone dioxide (SiO2). It is made by heating together sodium carbonate, calcium oxide (lime) or calcium carbonate (limestone) and silicone dioxide in the form of white sand. Sodium carbonate and calcium carbonate decompose when heated to give Na2O and CaO. These oxides combine with SiO2 in reactions to give CaSiCO3 and Na2SiO3, which exemplify the reaction of an acidic oxide, SiO2, with a basic oxide, CaO or Na2O, to give a salt, CaSiO3 or Na2SiO3. Numerous other oxides can be added to glass to give it special properties for use properties.19 The chemical composition of the GSs that we used in this experiment was similar to that of ordinary glass except for the addition of some oxides of metals, such as iron and aluminum, to increase mechanical resistance (table 1). Another advantage of glass is that it does not contain any components that are volatile in the body. For example, Aksu et al studied bone wax as an injection material but bone wax contains n-butyl acetate with a potential to metabolize into carbon dioxide and acetic acid in the body.18 In that aspect glass appears to be a safer material in respect to the elements of its composition. GS paste injected into the submucosa of the rabbit bladder preserved the original volume and position created by the first injection at month 1 and year 1 (fig. 3). Most nonautologous materials require some other substances to blend with local tissue to prevent dislocation. We think that since the GSs were greater than 150 ␮, they did not displace even in the rectus muscle at year 1. For example, dextronomer microspheres with a size of 80 to 120 ␮m are more prone to displace without the dextran polysaccharide vehicle.16 Similarly inert GSs do not lose volume at the end of year 1 regardless of the injection site (submucosa or rectus muscle)

1285

(fig. 3). However, many materials have been reported to fail to preserve their first injected volume, such as collagen.4 Another advantage of the larger size of GSs is the lack of distant migration risk. At the end of year 1 harvested organs such as the brain, lung, liver, kidney, spleen and pelvic lymph nodes did not show any migration. However, we studied only sample tissues and some sections histopathologically since histopathological study of the whole tissue is impractical, as previously reported.17 Since investigators have suggested that particle size smaller than 80 ␮ increases the migration tendency,17 we theoretically do not expect GS migration. PTFE particles are about 40 ␮ with a high risk of migration. In previous experiments PTFE particles were also found outside of the fibrous capsule.17 In sheep and rabbits regardless of the injection site we did not observe any sphere outside of the fibrous capsule that developed (figs. 2 and 4). GSs have already been used as an injection material in the literature with a smaller sphere size and with the ability to bond to soft tissues. As a paste (Bioglass), they were suspended in a viscous polysaccharide solution of sodium hyaluronate.20 Similarly bioglass was injected into rabbit bladder wall and subcutaneous tissue. However, only 67% nodule formation was observed. In our study after successful injection into the submucosa we always achieved a visible and durable nodule (fig. 3). We believe that GS size is the main determining factor of durable nodule formation. Another difficulty with bioglass is difficult injection.20 Nevertheless, we did not note any difficulty in injecting the material through an 18 gauge PTFE needle by a PTFE injection gun. Although bioglass seemed to be a good candidate for further animal studies, the report showing difficult injectability has hampered more investigations.20 Since we easily injected our paste using a PTFE injection setup, we advanced the experiment to testing in VUR animal models. The proper injection technique of GS paste into the submucosa stopped reflux in 13 of 15 refluxing ureters. An internal control was done by injecting vehicle into the other ureteral orifice and later by re-injection of GSs to correct VUR (table 2, fig. 5). This result supports the hypothesis of GS use as an injection material. We regret that we did not follow some sheep after unroofing without any injection to observe the natural course of unroofing to decide whether untreated reflux subsides. However, the correction of continuous reflux in vehicle-only injected ureteral units by GS injection seems convincing. Financial issues about injectable materials are also a major concern. It is imperative to develop an affordable material in low income countries. In investigations for the ideal injection material a particular effort should be made to decrease the cost of the invented material for widespread use in the world. In the financial aspect GSs promise an inexpensive alternative with a cost of less than $5. We believe that the biocompability, feasibility and efficacy studies of GS injection for the urinary tract encourages further study. CONCLUSIONS

Since the initial urological application of PTFE for the endoscopic correction of VUR, many novel injectable substances have been discovered and used. Each newly developed substance is generally more expensive than the older one. GSs can be an alternative to other injectable materials with lower cost, and similar efficacy and safety. Dr. Laurence S. Baskin reviewed the manuscript. REFERENCES

1. Stenberg, A., Hensle, T. W. and Lackgren, G.: Vesicoureteral reflux: a new treatment algorithm. Curr Urol Rep, 3: 107, 2002 2. Kershen, R. T. and Atala, A.: New advances in injectable therapies for the treatment of incontinence and vesicoureteral reflux. Urol Clin North Am, 26: 81, 1999

1286

GLASS SPHERES AS INJECTABLE MATERIAL FOR VESICOURETERAL REFLUX

3. Simsek, F., Dillioglugil, O., Ilker, Y. and Akdas, A.: Correction of primary and secondary vesicoureteral reflux by subureteric Teflon injection. Int Urol Nephrol, 27: 51, 1995 4. Frankenschmidt, A., Katzenwadel, A., Zimmerhackl, L. B. and Sommerkamp, H.: Endoscopic treatment of reflux by subureteric collagen injection: critical review of 5 years’ experience. J Endourol, 11: 343, 1997 5. Al-Hunayan, A. A., Kehinde, E. O., Elsalam, M. A. and Al-Mukhtar, R. S.: Outcome of endoscopic treatment for vesicoureteral reflux in children using polydimethylsiloxane. J Urol, 168: 2181, 2002 6. Kirsch, A. J., Perez-Brayfield, M. R. and Scherz, H. C.: Minimally invasive treatment of vesicoureteral reflux with endoscopic injection of dextranomer/hyaluronic acid copolymer: the Children’s Hospitals of Atlanta experience. J Urol, 170: 211, 2003 7. Atala, A., Peters, C. A., Retik, A. B. and Mandell, J.: Endoscopic treatment of vesicoureteral reflux with a self-detachable balloon system. J Urol, 148: 724, 1992 8. Zhang, Y. Y. and Bailey, R. R.: Treatment of vesicoureteric reflux in a sheep model using subureteric injection of cultured fetalbladder tissue. Pediatr Surg Int, 13: 32, 1998 9. Taneli, C., Ozcan, C., Ozdemir, N. and Gokdemir, A.: Correction of vesico-ureteric reflux by subureteric fibrin injection in dogs. Br J Urol, 74: 710, 1994 10. Chancellor, M. B., Rivas, D. A., Liberman, S. N., Moore, J., Jr. and Staas, W. E., Jr.: Cystoscopic autogenous fat injection treatment of vesicoureteral reflux in spinal cord injury. J Am Paraplegia Soc, 17: 50, 1994 11. Cendron, M., DeVore, D. P., Connolly, R., Sant, G. R., Ucci, A., Calahan, R. et al: The biological behavior of autologous colla-

gen injected into the rabbit bladder. J Urol, 154: 808, 1995 12. Atala, A., Cima, L. G., Kim, W., Paige, K. T., Vacanti, J. P., Retik, A. B. et al: Injectable alginate seeded with chondrocytes as a potential treatment for vesicoureteral reflux. J Urol, 150: 745, 1993 13. Dewan, P. A., Owen, A. J. and Byard, R. W.: Long-term histological response to subcutaneously injected Polytef and Bioplastique in a rat model. Br J Urol, 76: 161, 1995 14. O’Donnell, B. and Puri, P.: Treatment of vesicoureteric reflux by endoscopic injection of Teflon. Br Med J, 289: 7, 1984 15. Henly, D. R., Barrett, D. M., Weiland, T. L., O’Connor, M. K., Malizia, A. A. and Wein, A. J.: Particulate silicone for use in periurethral injections: local tissue effects and search for migration. J Urol, 153: 2039, 1995 16. Stenberg, A., Larsson, E. and La¨ ckgren, G.: Endoscopic treatment with dextranomer-hyaluronic acid for vesicoureteral reflux: histological findings. J Urol, 169: 1109, 2003 17. Malizia, A. A., Jr., Reiman, H. M., Myers, R. P., Sande, J. R., Barham, S. S., Benson, R. C., Jr. et al: Migration and granulomatous reaction after periurethral injection of polytef (Teflon). JAMA, 251: 3277, 1984 18. Aksu, K., Asci, R., Sarikaya, S., Buyukalpelli, R., Yilmaz, A. F., Yildiz, L. et al: Is Bone-wax an injectable urologic material? Eur Urol, 40: 564, 2001 19. Griffin, S. G. and Hill, R. G.: Influence of glass composition on the properties of glass polyalkenoate cements. Part II: influence of phosphate content. Biomaterials, 21: 399, 2000 20. Walker, R. D., Wilson, J. and Clark, A. E.: Injectable bioglass as a potential substitute for injectable polytetrafluoroethylene. J Urol, 148: 645, 1992