The Interaction of Urinary Components with Biomaterials in the Urinary Tract: Ureteral Stent Discoloration

Introduction

Urine is a complex medium consisting of 95% water, dissolved ions and salts, and organic compounds such as proteins, hormones, and metabolites. This complexity, compounded by the interindividual variability, significantly complicates the development of biocompatible materials for urinary devices. Given the highly variable chemical and physical characteristics of urine components, their interaction with the surface of indwelling urinary biomaterials involves various mechanisms, making the prevention of component deposition extremely difficult. While conditioning film deposition significantly changes the physical characteristics of device materials, chemical reactions between urinary molecules and device materials have remained understudied despite the significant potential for such reactions to change material composition and ultimately device performance. ^1,2

Ureteral stents are prone to complications including encrustation and bacterial adhesion/colonization, likely driven by the deposition of urinary components on the indwelling device surface. ^3,4 Recently, reports have emerged that discuss the “blackening” of indwelling ureteral stents, suggesting an association with increased encrustation and bacterial adhesion, leading to material changes and device failure. ^5,6 Given the implications of these studies with regard to potential negative effects of stent discoloration on stent function and the limited knowledge around mechanisms that may drive this phenomenon, we undertook the current work to better understand potential reactions that result in discoloration of some indwelling stents and determine whether this leads to material changes that increase device encrustation and bacterial adhesion, potentially compromising device performance. These studies are critical to determine whether colorimetric changes are cosmetic in nature or have potential compromising effects on patient treatment.

Methods

Following institutional ethics approval, 20 Polaris Ultra and 20 Percuflex Plus stents (same polymer) with varying degrees of discoloration were collected from patients stented for varying underlying medical conditions in the Department of Urology, Kitasato University School of Medicine, and The Stone Centre at Vancouver General Hospital. We did not focus on patients with the same medical condition as no single condition is associated with stent discoloration. Stents indwelling >48 hours were cystoscopically removed at time of routine removal and visually categorized by the same person based on depth of color into “dark” (black discoloration), “medium” (brown discoloration), and “light” (beige) discoloration. Due to discoloration not being uniform along the device length, samples were taken from proximal, mid, and distal sections and stored at −80°C. Stent pieces used for bacterial adhesion studies were washed in sterile artificial urine (AU) ⁷ to remove nonadherent material and immediately incubated with uropathogens. Analyses were conducted at The Stone Centre at Vancouver General Hospital Laboratories to limit variability in methodology.

Results

Surface characterization of control stent pieces (non-indwelling) via SEM/EDX revealed the presence of bismuth, used as a subcarbonate salt to increase radiopacity (Fig. 1). In contrast, discolored patient samples contained areas high in calcium and oxygen, typical for calcium oxalate deposits, high in sodium, carbon, chlorine, and phosphorous, typical for sodium chloride, calcium carbonate, calcium phosphate, and uric acid deposits (Fig. 2). In addition, discolored stents contained high levels of sulfur, leading us to pursue the hypothesis that stent discoloration may be the result of bismuth sulfide (H₂S) formation, known to be a black powder.

OPEN IN VIEWER

FIG. 1.

Gold sputtered scanning electron microscopy with energy-dispersive X-ray spectroscopy identified bismuth-containing polyurethane compound in the material of a control, non-discolored stent. Salt deposits show higher bismuth content than surface material.

OPEN IN VIEWER

FIG. 2.

Gold sputtered scanning electron microscopy with energy-dispersive X-ray spectroscopy identified bismuth-containing polyurethane compound in the material of a stent from a patient. Absorption spectra for three areas of the stent pictured are compared: sulfur detected, bismuth-containing polyurethane with a low concentration of nitrogen, high concentration of calcium and oxalate.

In vitro stent discoloration

To validate a potential chemical mechanism leading to discoloration, stent pieces were incubated in 1% sodium sulfide as a sulfur donor and increasing concentrations of HCl (0, 0.02, 0.05, 0.075, 0.1 M) as the acid needed to produce H₂S. Overall, increasing concentrations of HCl resulted in a similar color change to that seen on clinical samples. Specifically, higher concentrations (0.075 and 0.1 M HCl) of the acid resulted in more rapid and distinct discoloration (black color change) compared with lower concentrations (0.02 and 0.05 M HCl), which resulted in more subtle color changes (brown/light gray color change) (Fig. 3).

OPEN IN VIEWER

FIG. 3.

Stent pieces in 1% sodium sulfide with varying concentrations of a reducing agent, HCl: (A) 0 M HCl, (B) 0.02 M HCl, (C) 0.05 M HCl, (D) 0.075 M HCl, (E) 0.1 M HCl. HCl = hydrochloric acid.

Stent encrustation

To determine whether stent discoloration increases its encrustation potential, we first studied the encrustation of stent pieces discolored to varying degrees using our in vitro model (more controlled discoloration) to those of non-discolored control material. Overall, stent pieces of increasing discoloration did not show statistically significant differences in calcium encrustation (p = 0.169, Fig. 4A) compared with less discolored/non-discolored stent pieces. To verify this for clinical samples, we compared the encrustation of variably discolored clinical stent pieces from both sites and also found no statistically significant differences in the degree of encrustation of: (i) discolored stents from Japanese vs Canadian patients when compared with each other or control stents (p = 0.520, Fig. 4B) or (ii) between non-discolored patient stents, discolored patient stents, and control stents (p = 0.552, Fig. 4C). Interestingly, there was a nonsignificant trend toward lower levels of calcium encrustation in discolored patient stents from both countries compared with non-indwelling control stents.

OPEN IN VIEWER

FIG. 4.

(A) Calcium encrustation levels on in vitro discolored stent pieces with 1% sodium sulfide and increasing concentrations of HCl. Following in vitro discoloration, stent pieces were incubated in artificial urine for 3 weeks. (B) Calcium encrustation levels on control stents and stents pieces from patients from Japan and Canada. Stent pieces were incubated in artificial urine for 3 weeks. (C) Calcium encrustation levels on control stents and non-discolored vs discolored patient stents from Japan and Canada. Stent pieces were incubated in artificial urine for 3 weeks. Crystals on stents were dissolved with lanthanum chloride solution, and atomic absorption spectroscopy was used to measure the concentration of calcium in solution.

Bacterial adhesion to discolored patient and control stents

To test whether stent discoloration triggers material surface changes that increase bacterial:surface interactions, we tested the adhesion of the uropathogen S. aureus to varying sections of discolored patient stents (bladder curl, bladder end straight section, middle of the stent, kidney end straight section, and kidney curl) and compared them with that on representative non-discolored control stent pieces. Different stent sections were chosen to ensure that adhesion was tested along the entire stent length, allowing us to identify potential differences in adhesion to different areas of the stent based on varying physical/chemical characteristics including degree of discoloration. Overall, we did not find any significant difference in the number of adherent bacteria on different segments from discolored stents compared with non-discolored control stents (p = 0.2714, Fig. 5). Comparing bacterial adhesion between discolored patient stents and control stents revealed no significant differences in mean CFU/mL between different stent sections or differently discolored stents (p-value). Furthermore, no significant differences in bacterial adhesion were observed between variably discolored stent segments on the same stent, as no one stent was ever found to be completely discolored.

OPEN IN VIEWER

FIG. 5.

(A) Mean CFU counts for Staphylococcus aureus adhesion to control stent pieces and discolored stent pieces grouped by section of the stent. (B–E) Bacterial adhesion and colonization to stent pieces following increasing in vitro discoloration. (B) Adhesion of uropathogenic Escherichia coli C1214 (4 hours), (C) colonization of uropathogenic E. coli C1214 (48 hours), (D) adhesion of S. aureus (MRSA) (4 hours), (E) colonization of S. aureus (MRSA) (48 hours). Overall, no significant differences were seen in the adhesion or colonization of either uropathogen with increasing stent discoloration. CFU = colony-forming unit; MRSA = methicillin-resistant S. aureus.

FTIR microscopy analysis of material surface

To determine whether stent discoloration is associated with unusual encrustation on the material surface or a change in stent material composition, we performed FTIR microscopy analysis of variably discolored stent pieces removed from Canadian (n = 5 total) and Japanese (n = 7 total) patients. Overall, the FTIR spectra fell into one of four absorption spectra differentiated by the presence or absence of a small number of bands (Fig. 6). Stent pieces in Group 1 showed a similar absorption pattern to the non-discolored control stent, with the exception of no signal at 3361 cm⁻¹ and 1566 cm⁻¹. Stent pieces in Group 2 again had similar absorption spectra compared with the control stent; however, the stent pieces did exhibit a shifted band at 3292.5 cm⁻¹ and the presence of strong bands at 1642 cm⁻¹ and 1546 cm⁻¹. Similarly, stent pieces in Group 3 had a shifted band at 3283.3 cm⁻¹ and the appearance of band at 1652 cm⁻¹ and 1539.3 cm⁻¹. Finally, stent pieces in Group 4 had weak shifted signals around 3292 cm⁻¹ and a band appear at 1300 cm⁻¹.

OPEN IN VIEWER

FIG. 6.

Seven stents from Japanese patients and five stents from Canadian patients were categorized into four groups based on similarity of Fourier transmission infrared spectroscopy absorption peaks. Stents were grouped as follows: (A) one dark Japan, (B) one dark Japan, one medium Japan, one dark Canada, one medium Canada, one light Canada, (C) one light Japan, one light Canada, one medium Canada. (D) One medium Japan, one light Japan, one dark Canada.

Despite each group being characterized by a specific absorption pattern, none of these differed significantly from that obtained using non-indwelling non-discolored control stent pieces.

Discussion

Urine is a complex medium, with highly variable composition and comprising varying chemistries. Given the constant exposure of indwelling devices to urine, direct interactions between the medium and biomaterials are inevitable, generally manifesting as a layer of organic material and encrustations. ² While these interactions and material surface changes have been accepted throughout urology, most recently reported incidents of severe stent discoloration have resulted in speculation that this phenomenon is associated with increased encrustation and bacterial biofilm formation, significantly increasing the risk for additional complications due to stent failure. ^5,6 Given the potential impact of such speculation on the future use of ureteral stents, improving our understanding of the root cause for stent discoloration and its potential impact on stent function are warranted.

Aside from the polymer material itself, ureteral stents tend to have additives required to make the otherwise translucent stent radiopaque. These include bismuth subcarbonate or barium sulfate. Our initial SEM/XRD analysis identifying the presence of barium in Polaris/Percuflex non-discolored control stent material led us to investigate the possibility that the dark discoloration observed clinically could be due to a chemical reaction that converted the colorless bismuth subcarbonate stents to the black bismuth sulfide, a common reaction observed in the stool of patients taking bismuth-based antacids such as Pepto-Bismol. Subsequent in vitro studies incubating non-discolored stent pieces in the presence of sodium sulfide (sulfur donor) and HCl as the acid needed to generate H₂S confirmed this hypothesis, resulting in rapid discoloration of the stent material, with the acid concentration determining the degree of discoloration. This suggests that at least in part, pH may drive the discoloration process. That said, since the potential chemical reaction leading to stent discoloration requires several components, urine pH alone cannot be the trigger.

Urine is a complex medium containing numerous potential sulfur and proton donors; however, considering that severe stent discoloration is not a common clinical occurrence, these are unlikely to be abundant and obvious. Given this, their identification would be very complex, especially given that they are likely to vary between patients. It must be pointed out that our in vitro studies triggering stent discoloration do not serve as definitive proof that the specific components used are what trigger the color change in patients, but rather serves to validate a hypothesized mechanism based on components within the stent material, that a reaction between bismuth subcarbonate and a sulfur donor in the presence of an acid leads to stent discoloration similar to that observed clinically. Definitive proof of this will require delving into potentially complex investigations to identify differences in urinary factors between patients with severe stent discoloration and those without and is beyond the scope of this work.

To answer the question whether stent discoloration leads to clinically relevant changes, subsequent analyses were carried out on clinical samples. FTIR microscopy analysis of stent pieces of varying degrees of discoloration removed from patients in both Vancouver and Japan revealed no significant differences in spectra obtained from either sample indicating that the bulk surface material of discolored stents is the same as that of a non-discolored control stent.

While we did notice slight changes in the spectra around 3361 cm⁻¹ and 1556 cm⁻¹ and appearance of bands at 1652 cm⁻¹ and 1539.3 cm⁻¹ in discolored stents, these are likely attributed to the loss of the hydrophilic coating found on stents to increased lubricity upon insertion rather than the presence of encrustation or change in material. This coating is lost shortly after stent insertion, making this hypothesis plausible. Collectively, these results indicate that the color change of ureteral stents is unlikely to be associated with the deposition of an unusual conditioning film/encrustation or an overall change in the actual stent material, but rather is cosmetic in nature.

Previous work suggested that stent discoloration is associated with increased material encrustation and bacterial adhesion, significantly increasing the risk for stent failure. While those studies did not include specific molecular analyses to support this, the present molecular-based work did not observe any significant differences in the degree of calcium-based encrustation or bacterial adhesion and colonization of in vitro discolored stents (as a positive control of discoloration) or actual stents removed from patients by both gram-positive and gram-negative species. It must be pointed out that the models used to test fouling of the stent material (both patient and in vitro samples) by crystal deposition and bacterial attachment were carried out using conditions that favor their deposition due to a supersaturated AU and exposure to a significantly higher number of bacteria than would be encountered in vivo. The fact that these stringent conditions did not result in increased fouling of the discolored material surface strongly suggests that discoloration does not promote increased device encrustation and/or bacterial colonization.

Conclusion

A potential mechanism driving stent discoloration may involve the interaction of urine-derived sulfide with bismuth disulfide added to stent materials to increase radiopacity, producing the darker colored bismuth sulfate. Extensive subsequent characterization of in vitro and clinically discolored stents indicated that discoloration is not associated with significant changes in stent material, increased encrustation, or bacterial adhesion. Stent discoloration is therefore believed to be a cosmetic change in stents containing bismuth subcarbonate as the radiopacifying agent and is unlikely to have clinical consequences.