EUO prevalence is growing steadily and continues to be a major treatment challenge for the urologist . Different stent types and configurations have been reported in an effort to adequately drain kidneys and prevent complications. Over recent years, tandem ureteral stents [4,5,6,7,8,9], metal stents [10,11,12,13,14,15] or metal-mesh stents [16, 17] have been advocated as the preferred drainage method under EUO. However, few “head-to-head” studies comparing all of these options are available, which can be attributed to limitations that include relatively low numbers of patients with EUO that can be recruited at a single center, different etiologies responsible for EUO, and differences in ideal stent preferences between medical centers. We present here the first comparison of a wide variety of different ureteral stents subjected to EUO in an in vitro model of a stented ureter. Our experiments indicate that neither tandem ureteral stents nor metal stents offer any clear advantage over large luminal polymeric stents in preventing stent failure, although they appear more effective than single, small luminal polymeric stents (comparing failure times to those reported in ). In addition, endopyelotomy stents show no superiority over large luminal polymeric stents under EUO conditions of deformation, external ureteral compression and fluids containing colloids.
In a retrospective study, Askawa et al.  compared polymeric stents to Resonance® stents in patients with EUO. A total of 54 ureters in 35 patients were drained with polymeric stents, and 72 ureters in 57 patients were treated with Resonance® stents. Overall stent patency for both groups was 70%, while Resonance® stents showed higher patency rates when compared to polymeric stents after one year of follow-up (78% vs. 61%). This difference did not reach statistical significance in multivariate analysis. The authors  concluded that the Resonance® stent is superior to polymeric stents as it better resists external compression forces, so that it should be the first choice to drain EUO. Another comparative study by Liu et al.  evaluated prospectively, in a non-randomized fashion, 104 patients with EUO using single or tandem 7F polymeric stents based on patient preference. Stent patency duration among 63 renal units drained with tandem stents was 214 days, in comparison to 176 days among 94 patients with single stents. The authors concluded that tandem polymeric stents are superior to single stents.
These authors [7, 13] suggested that a key etiology responsible for stent failure under EUO is external compression forces exerted over the ureter, resulting in deformation, kinking and closure of the stent and ureter lumina. While it may be intuitive to expect that polymeric stents show poor resistance to external ureteral pressure, we showed previously [18, 21] that this is not the case via in vitro experiments. Polymeric stents (4.8F to 8F) were subjected to both ureteral deformation and increasing external forces (compression) of up to 5000 g; the 4.8F stent failed at high compression, while the 6F, 7F and 8F stents remained patent throughout. We concluded [18, 21] that only unrealistic external pressure forces may lead to stent obstruction and therefore external pressure/deformation as sole etiology for stent failure is less reasonable. Two biases should be noted in the two above-mentioned studies [7, 13] comparing polymeric stents to either tandem or metal stents. First, the luminal size of the polymeric stents that were compared to the Resonance® stent was not mentioned; use of a relatively small luminal size might have influenced the results. Second, the tandem and single stents in the comparative study  were all 7F. Clearly, the likelihood for a stent to remain patent is higher when comparing two stents of the same size in one ureter to a different ureter drained with only one stent of the same size. A valuable comparison would have been between 7F tandem stents and a single 8F polymeric stent.
In the current experiment, only the 8F stent remained patent consistently throughout the experiment duration, while all other stent configurations—tandem 6F and 7F, Resonance®, endopyelotomy—generally failed. The exact mechanism in which ureteral stents fail under EUO is unclear, although several possible risk factors have been suggested [2, 7], including stent incrustation/encrustation due to mineralization, blockage by debris accumulation, loss of ureteral peristalsis, tissue growth through stent side holes and/or stent buckling. As described previously , colloidal fluid can play a critical synergetic role with deformation and compression in the occurrence of stent failure. Assessment of different luminal polymeric sizes stents subjected to EUO in an in vitro model resulted in stent failure of the smaller luminal sizes (4.8F and 6F) as opposed to the larger sizes (7F and 8F) that remained open. We note, too, that tandem stents have been reported to be effective in some clinical settings under EUO [5, 7, 9, 22]. We speculate that this may be due to movement between the stents, which may reduce colloid aggregation and accumulation of colloids, as well as to the possibility that tandem stents maintain a space between the stents to allow urine flow even subject to ureter and stent lumina obstruction.
The time to failure variation of each stent type and configuration is likely related to the “random” nature of colloid accumulation over time. In many cases, we noted steady rising renal pressure and then sudden pressure drops, which were most likely due to accumulation and subsequent release of colloids. Whether mobilization of patients with EUO assists in prevention of colloid accumulation has yet to be proved. We emphasize that we make no claim regarding actual times to stent failure in clinical settings, nor do we consider EUO progression over time. Rather, the experiment was designed to work with a specific colloidal concentration and examine the relative times to failure (or not) among the different stent sizes and configurations.
Existing literature and the results reported here indicate that larger-lumen stents are less likely to become occluded with debris, due to the higher flow rates within them. Because urine flow in a ureter-stent system is generally laminar , illustrative flow calculations can be derived from Poiseuille’s law. We note, too, that while the relative contributions of stent and ureter lumina to overall flow remain poorly understood, the stent lumen likely controls flow behavior particularly in the vicinity of the EUO where the ureter lumen is obstructed, which ultimately affects overall flow through the ureter-stent system; we therefore focus on stent flow in this region. The volumetric flow, Q, through a tube is Q = (πPr4)/(8ηl), where the dynamic viscosity for urine is  η ≈ 8.5 × 10−4 Pa·s at 37 °C and r is the internal radius. We assume a straight stent length l = 24 cm (no pigtails), and total stent radii of r = 0.8, 1.0, 1.17 and 1.33 mm (corresponding to 4.8F, 6F, 7F and 8F stents, respectively). We modify r to account for wall thickness (measured  as 0.22 mm for the 4.8F, 0.4 mm for 6F, 7F, 8F stents), and impose a renal unit pressure of 1 cmH2O (P = 98.0665 Pa). Calculations of Q in different stent configurations are given in Table 2. Notably, increasing stent luminal diameter, e.g., from 6 to 8F, increases volumetric stent flow by a factor of 5.9. In this context, consideration of tandem stents does not necessarily lead to increased volumetric flow rates. For example, tandem 6F stents have an effective, combined cross-sectional area of 6.28 mm2, while a single 8F stent has an effective cross-sectional area of 5.55 mm2; and yet, the volumetric flow in a single 8F stent is still a factor of 2.9 higher than in tandem 6F stents (and a factor of 1.1 higher than tandem 7F stents). These results are controlled by the strong effects of frictional forces on the stent walls, which vary significantly as a function of radius—critically, Q varies with the factor r4.
The limitations of our study are related to the model design [18, 21]; see Additional file 1 for discussion regarding latex tubing, EUO shape and pressure, and type of colloidal solution (rather than “artificial urine”). Clearly, an in vitro experimental study cannot account for all physiological properties in the human in vivo environment. As such, the findings reported here do not correlate directly to clinical findings, particularly in terms of times to stent failure. However, our systematic analysis among ureter-stent configurations yields important relative comparisons of their failure dynamics. The latex tubing simulating a ureter has physical characteristics different than a real ureteral wall, and the lack of (even minimal) peristaltic motion may have some effect on the results (although we note that loss of ureteral peristalsis occurs frequently with stenting [2, 7, 24]). Moreover, dynamic ureteral responses caused by indwelling stents and EUO will likely further affect the flow dynamics over time, such as reflux or retrograde pressure transmission through and/or around the stent. The results here also are a function of the type and concentration of colloid material, and the use of solution containing no other salts, enzymes and organic material. As such, stent mineralization, various colloid aggregation properties, and biological activity remain unaccounted for. Moreover, the occurrence of blockage and the actual times to blockage (and variability among replicate experiments) are also affected by colloid concentration. Notwithstanding the above, this is the first study to systematically measure effects of colloidal fluid flow in a comparative study of single, tandem, metal, and endopyelotomy stents in stented-ureter systems, both under and without EUO.