Endoscopic injection therapy for treatment of vesicoureteric reflux: A 20-year perspective

Abstract

OBJECTIVE:

To review the application and outcome of endoscopic injection therapy for vesicoureteric reflux in regard to its evolution over the past two decades.

DATA SOURCES:

Review articles, original reports and abstracts pertaining to endoscopic injection therapy were obtained through a PubMed search of English, German and French publications from 1981 to 2001.

DATA SELECTION:

A total of 46 studies were selected. Four were selected to support basic concepts in the management of vesicoureteric reflux, and the remainder pertained specifically to endoscopic injection therapy for vesicoureteric reflux.

DATA EXTRACTION:

The reports were analyzed with focus on the physical properties of the biomaterial injected, results of treatment in regard to the cure of vesicoureteric reflux, duration of cure, and possible adverse effects and clinical benefits engendered by the use of injectable materials.

DATA SYNTHESIS:

Endoscopic injection therapy successfully cures vesicoureteric reflux in 60% to 80% of cases. Success rates are higher with particulate materials (Teflon and Macroplastique) than with bovine collagen or autologous chondrocytes. Long term data regarding cure are scant. Although concerns about particulate migration and autoimmune disease exist, these have not been borne out of clinical experience. Endoscopic injection may be accomplished on an outpatient basis, with less morbidity than with open ureteroneocystostomy.

CONCLUSIONS:

Endoscopic injection therapy should be offered as an alternative treatment in patients with indications to consider ureteroneocystotomy, but should not change the indications for surgical intervention. The ideal biomaterial for injection has yet to be developed, but the field of autologous tissue engineering holds promise for future development.

Paediatricians commonly attend to infants and children with urinary tract infections (UTIs), and in such a cohort, vesicoureteric reflux (VUR) may be found in as many as 35% (1). The occurrence of pyelonephritis is facilitated by the presence of VUR, and the end result may be renal scarring, which may produce long term morbidity in the form of hypertension and renal insufficiency (2). Because sterile reflux is not believed to produce renal damage, and VUR has a propensity for spontaneous resolution, most children can be managed expectantly on suppressive antibiotics (3). There are children who require surgical intervention for the management of VUR, particularly those who have documented febrile breakthrough UTIs. For years, ureteroneocystostomy has been the surgical ‘gold standard’ in the management of such patients. With cure rates of approximately 98%, this technique is a proven intervention (4). Unfortunately, ureteroneocystotomy is often associated with morbidity, comprising primarily bladder spasms or voiding dysfunction, and is associated with a variable length of inpatient hospitalization. In 1981, Matouschek (5) first described the technique of endoscopic injection of Teflon (DuPont, USA) paste under the ureteric orifice to correct reflux. Since that time the children treated with this outpatient technique numbers in the thousands and the materials used for this procedure have evolved. The purpose of the present manuscript is to review the technique of endoscopic injection, and the pros and cons of the varied materials that have been used for this indication. The focus will primarily be on the results of treatment and the potential risks inherent in the use of such biomaterials. Hopefully, this review will crystallize the usefulness and limitations of endoscopic injection for the practising paediatrician.

MATERIALS AND METHODS

The endoscopic subureteric injection technique for the treatment of VUR has become known as the ‘STING’ procedure in the urological vernacular. While initially coined by O’Donnell and Puri (6) to describe the ‘subtrigonal ingection’ of Teflon paste, STING is a term now widely applied, regardless of the injectable material used. This procedure was conceptualized as an outpatient technique, and in the majority of cases, this may be accomplished. Under a general anesthetic and with parenteral antibiotic prophylaxis, the patient is prepared and draped in the modified lithotomy position. Cystoscopy is then performed with an appropriately sized instrument, and an injection needle is advanced through the operating channel of the cystoscope to approach the ureteric orifice at the 6 o’clock position. The needle is advanced in the plane between the bladder mucosa and bladder muscle, starting 4 mm distal to the ureteric orifice and then travelling under the intravesical ureter for a distance of about 5 mm. The injection is performed until the appearance of the ureteric orifice resembles an inverted crescent on a hillock of injected material (Figure 1). The needle remains in place for 1 to 2 min to prevent any back leak of the injectable, and the instrument and needle are then removed. It is very important to make the first needle placement accurately because multiple needle punctures under a given ureteric orifice lead to back leak of the material with resultant failure. The volume of the injection depends on the severity of the defect at the ureterovesical junction, the accuracy of needle placement, and the type of material injected, but generally ranges from 0.1 to 2 mL. The patient is discharged home the same day on oral antibiotic prophylaxis, with plans for a renal ultrasound in four to six weeks and a cystogram three months after treatment.

Figure 1).

a The endoscopic injection needle approaches the 6 o’clock position of the ureteric orifice and pierces the mucosa approximately 4 mm distal to the orifice. The needle is then advanced for approximately 5 mm in the plane between the ureter and the bladder muscle. b The implant is then injected until the ureteric orifice resembles an inverted crescent sitting on a hillock of the injected material. c The implant is seen in the ideal location within the submucosal space, producing coaptation of the ureteric orifice

The quest for the ideal injectable material continues, despite two decades of experience with STING. The ideal injectable material should have anatomic integrity, as evidenced by the ease of endoscopic delivery and maintenance of implant volume over time. It should also demonstrate biological safety in the sense that it should be biocompatible, nonmigratory, nonantigenic and noncarcinogenic. In general terms, injectable materials may be nonautologous or autologous. The nonautologous materials that have been used in children are polytetrafluoroethylene paste (Teflon or polytef), glutaraldehyde cross-linked bovine dermal collagen (Zyplast [McGhan Medical Corporation, USA] or Contigen, [CR Bard Inc, USA]), polydimethylsiloxane (Macroplastique [Uroplasty, The Netherlands]) and dextranomer microspheres in sodium hyaluronan solution (Deflux [Q-Med, Toronto]). The autologous materials that have found applications are fat, blood and chondrocytes. These materials are discussed in regard to their clinical track record and current status.

RESULTS

Nonautologous materials

Polytetrafluoroethylene (Teflon or polytef) paste

Teflon paste was the first material investigated for the endoscopic treatment of VUR (5,6). This paste is comprised of an equal mixture of polytetrafluoroethylene (polytef) particles in glycerine. After injection, the glycerine is absorbed, and the polytef is retained within a fibrous capsule (7). The first substantial clinical series was reported in 1984 by O’Donnell and Puri (6) from Dublin. Subsequently, many other centres have reported their experiences with polytef injection for reflux, and long term data have recently been made available from the Dublin group. If one enumerates reports of 4234 patients (6316 ureters) treated with polytef for VUR, the success rate after a single injection approaches 75.9%, but additional injections may increase the cure rate to 84.9% (8,9). Long term data from one to nine years after treatment suggests a relapse rate of 7% (8). Thus, the technique and material appears to be promising, if not for the spectre of patient safety concerns. Malizia et al (10,11) first raised these concerns when they reported an experimental study in rabbits and dogs, in which the periurethral or subureteric injection of Teflon caused local granulomatous reactions characterized by fibrosis and foreign body giant cells. In addition, they documented a distant spread of Teflon particles to regional lymph nodes, the lungs and the central nervous system. Teflon particles vary in size between 4 and 40 μm, and such spreading is thought to be due to embolic phenomena. The implications of such experimental findings in humans are unknown. There exists a small number of clinical case reports in which Teflon has been documented to migrate distally after injection into the lower urinary tract (12,13). However, the large European experience with this technique suggests that these concerns are overblown (8). The widespread use of this material in North America has been curtailed by safety issues, while it remains in widespread use elsewhere in the world.

Glutaraldehyde cross-linked bovine dermal collagen (Zyplast or Contigen)

Injectable bovine collagen was the next material considered for use in the STING procedure. This material is extracted from bovine dermis by enzymatic digestion and then cross-linked by the addition of glutaraldehyde. When supplied as Zyplast or Contigen, it is sterile and nonpyrogenic. The cross-linking serves to stabilize the integrity of the implant over time and reduce its immunogenicity. After the implant is injected it becomes encapsulated and recruits host blood vessels and fibroblasts, acting as a scaffold for the host to lay down autologous collagen (14,15). The first large clinical series reporting the use of collagen in STING was that of Leonard and associates (16) from Johns Hopkins Hospital, Baltimore, Maryland. Subsequently, additional reports from Europe have added to the information available on this substance. If one enumerates the reported results, the success rate after one injection approaches 60% (16–18). Just as with polytef, repeat injections of collagen may increase the cure rate from 65% to 80% (16,17). The use of collagen for STING has come under scrutiny over two main issues: durability and potential immunogenicity. Regarding durability, several studies have documented that reflux relapse rates of 10% to 20% may occur with long term follow-up of patients undergoing collagen STING (17,19). Haferkamp et al (20) have reported recurrence rates as high as 80% when comparing results at three months postinjection to three years postinjection. However, they only offered their patients one injection. Improper implant placement, material shift or degradation with time may have accounted for their unfavourable experiences (21). The take home message is that collagen is likely not as durable as polytef in terms of long term efficacy. However, it may provide a permanent cure in many, and allow other patients to pass through tumultuous times in their clinical courses without the need for open surgery or suppressive antibiotics. The issue of the immunogenicity of collagen implanted in the urinary tract was initially addressed in adults undergoing large volumes of collagen injection around their bladder necks as treatment for urinary incontinence (22). A subsequent study performed in children undergoing smaller volume collagen implants during STING also concluded that approximately 30% of patients will develop serum antibovine collagen antibodies in response to the treatment (23). However, none of the patients developing antibovine collagen antibodies after STING had antibodies that cross-reacted with human collagen, nor did any have clinical events compatible with autoimmune disease. The issue remains controversial, as Cukier et al (24) reported a higher than expected incidence of dermatomyositis and polymyositis in adults who have undergone intradermal bovine dermal collagen injection for cosmetic indications. Collagen is approved by the Therapeutic Products Protectorate (TPP) for the application of bladder neck injection for urinary incontinence. The lack of approval for the indication of STING is not due to safety concerns, but rather due to packaging issues, in that the volume of the aliquots supplied for bladder neck injection far exceed the requirements for STING. TPP has approved STING on a case by case basis, provided the hospital pharmacy can repackage the collagen in smaller aliquots under sterile conditions, and the reports of patient progress and adverse events are filed with the agency at set intervals after treatment.

Polydimethylsiloxane (Macroplastique)

This material comprises particulate silicon microimplants (140 μm) of fully vulcanized polydimethylsiloxane suspended in a water-soluble polyvinylpyrrolidone hydrogel in a 60:40 ratio (25). Just as with other implants, Macroplastique becomes encapsulated and allows for the inward migration of host fibroblasts that deposit autologous collagen (26). The initial reports on the results of Macroplastique STING came from Europe, and claimed 80% to 90% success after one injection (27,28). Recent Canadian experience with Macroplastique STING has been reported from Toronto (29). Seventy-four children (111 renal units) were treated, and 76% of the children (81% of renal units) were cured after one injection. Repeat injection increased the cure rate to 84% of the children (90% of ureters). Follow-up of the treated patients was limited to 24 months and the cystogram was generally performed three months after the last injection (29). Although there is a paucity of long term follow-up data for STING, cures after Macroplastique injection for urinary incontinence have been documented for up to three years postinjection (30,31).

Just as with any other biomaterial, safety concerns exist for Macroplastique. In animal studies, silicone microparticles have been documented to migrate distally to the lungs, kidneys, brain and lymph nodes after periurethral injection (32). Regarding human data, a recent study from France (33) examined the histology of Macroplastique implants in patients who failed STING and went on to ureteric reimplantation. It was demonstrated that Macroplastique engendered a more intense local inflammatory reaction than Teflon, and was fragmented into 6 μm particles by macrophage action. Such particles are of a concern because they are small enough to allow for distant migration. Thus, it is apparent that safety concerns regarding this material echo those of Teflon. This material is approved by TPP for bladder neck injection and STING.

Dextranomer in sodium hyaluronan (Deflux)

Deflux consists of dextranomer microspheres of 80 to 120 μm in diameter in a 1% molecular weight sodium hyaluronan solution. The sodium hyaluronan serves as a vehicle and is absorbed by the body over the course of one to two weeks (34). In animal studies, the injected material engenders an inflammatory reaction with ingrowth of host vessels and fibroblasts, with the ultimate deposition of autologous collagen (35). Furthermore, animal studies have documented a lack of migration of the injected material from bladder to distant organs (36). The clinical data regarding the use of Deflux for the treatment of VUR emanates primarily from the Swedish group responsible for the initial experimental studies quoted above. They treated 75 patients (115 ureters) and achieved success at three months after one injection in 68% of the cases. They also noted that 16 of 18 ureters cured at three months were also reflux-free one year after treatment (34). Obviously, significant long term data are lacking. Currently there is a plan to start a North American based multicentre study to generate data with view to Food and Drug Administration (FDA) submission and approval of Deflux as a treatment for VUR.

Autologous materials

Fat

Autologous fat would seem to be an ideal injectable substance regarding its ease of procurement and ubiquity. Initial studies done on rat bladders documented that fat implants lost 25% to 50% of their volume with time, but did remain in situ up to 11 months after injection (37). However, clinical studies using autologous fat STING to correct VUR are limited, and the results are disappointing. Chancellor et al (38) documented that two of seven patients were cured of reflux six months after injection, while Palma et al (39) had only one cured ureter among 17 treated. Obviously, such results fail to match the success seen with other injectable materials, and have eliminated autologous fat from the options available for STING.

Blood

Autologous blood would seem to be another readily available source of material for STING. There exists only one English language paper describing its use (40). In this study, the patient’s own heparinized blood was injected subureterically. Thrombin and protamine were added before the needle was withdrawn. A total of 13 adult patients (16 renal units) were thus treated, with nine of 16 renal units cured at three months after one injection. Given the lack of further publications using autologous blood for this indication, it is unlikely that it will remain a viable alternative.

Autologous chondrocytes

With the advent of tissue engineering, the concept of harvesting a patient’s own tissue and growing it in vitro for subsequent autologous usage became feasible. Atala and colleagues (41) demonstrated the ability to harvest swine auricular cartilage, expand the harvested chondrocytes in vitro and then inject them in an alginate vehicle to treat VUR in these same swine. In this study, the VUR was cured up to six months after injection. Furthermore, at animal sacrifice, the injected implants were seen to be comprised of cartilage with minimal surrounding inflammation. Moreover, there was no evidence for distal migration of the chondrocyte implants. This experimental work generated excitement in urological circles, and finally led to the initiation of an FDA-approved clinical trial. To date, the trial has recruited 29 children (47 ureters). The cure rate at three months after one injection is 55% of ureters treated, which increased to 86% after a second or third injection. At one year after the last injection, cure is maintained in 70% of the ureters (65% of patients). Three patients underwent ureteroneocystostomy after failed injection, and the histology of the explanted chondrocyte injection revealed calcified alginate without evidence of viable chondrocytes (42). Unfortunately, these results are no better than those achieved with bovine collagen, and the patients must undergo an extra general anesthetic for cartilage harvest. It is disappointing that viable cartilage was not seen in the explants, but this may be the reason that they failed, while their counterpart who had viable cartilage over the long term may have been the ones with persistent cure. It could also be that the alginate is acting as the bulking agent, and the cartilage is an unnecessary component. Obviously, these results have generated some ‘food for thought’, and, while this concept should not be abandoned, clearly more work needs to be done to clarify these issues.

DISCUSSION

The concept of STING as a surgical intervention aimed at curing VUR is well established. With two decades of experience and thousands of patients treated, it remains an option when counselling patients and parents on the need for surgery, given the appropriate indications. The advantage of STING is that it is minimally invasive, may be accomplished as an outpatient procedure with minimal morbidity, and is more cost effective than open surgery (43,44). However, this must be balanced against the fact that STING has lower cure rates than open surgery; necessitates, in most cases, the injection of a foreign material, with potential for adverse reaction; and has limited data regarding long term cure rates. Given these potential limitations, one has to be careful regarding indications for intervention. Many investigators believe that because STING is a rather innocuous outpatient procedure, we should be more liberal in offering it to patients with VUR who would not normally come to surgical intervention. However, one must not lose sight of the fact that it is a surgical intervention. Thus, STING carries with it the risks of general anesthetic, bleeding and UTI, just as would an open reimplant. Furthermore, there have been instances of ureteric obstruction secondary to STING, and, given the limitations of the technique and its inherent learning curve, the failure of the intervention is comparatively high. If one were to recommend STING to a patient, which material should be used? It seems the best results are obtained with particulate materials such as Teflon and Macroplastique, but these are the materials in which local inflammatory reactions and distant migrations may be more problematic. Currently, bovine collagen is the only other alternative available in Canada, and although its cure rate is somewhat lower than that of the microparticulate injectables, it is thought to be safer. The future may yield more appropriate injectable materials, such as improved formulations of autologous collagen, autologous smooth muscle or detachable silicone balloons (45,46). However, ureteroneocystostomy remains the standard of care to which any new injectable material must measure up. The STING is here to stay, but its presence should not liberalize our indications for surgical intervention in patients with VUR. Moreover, we must be vigilant in following patients who have had STING regarding the long term cure rate and safety of the treatment.