Percutaneous Nephrostomy versus Retrograde Ureteral Stent for Management of Malignant Ureteral Obstruction in Adults: A Systematic Review of the Literature

Abstract

Background: Malignant ureteral obstruction (MUO) is a common presentation in advanced urological and non-urological malignancies. Percutaneous nephrostomy (PCN) and retrograde ureteral stent (RUS) are the most commonly performed procedures to relieve the obstruction. The comparative effectiveness of PCN and RUS for decompression of MUO remains uncertain. Purpose: To systematically review the literature for evidence of improved efficacy of one of these procedures in terms of renal function preservation and clinical outcomes. Methods: We searched Ovid Medline, Ovid EMBASE, CINAHL, Cochrane Central Register of Controlled Trials (CENTRAL), and Scopus from the date of inception to October 2022. In addition, gray literature was searched through OpenGray (https://opengrey.eu/), dissertation and thesis database (ProQuest) via (https://www.proquest.com), and Clinical trial.gov website. The reference lists of all the included studies were also searched. Two reviewers independently reviewed and selected studies, assessed the quality, and extracted the data. Results: Overall, 25 eligible studies including 1864 patients compared PCN and RUS (head-to-head). PCN and RUS were found to be similarly effective in improving renal function. However, PCN appears to be superior in maintaining this reduction. The complication rate and quality of life were comparable between the 2 methods, but the length of hospital stay and the financial cost were significantly higher in the PCN group. The mean technical success rate in RUS was 70.3% (21% to 100%) and in PCN was 98.8% (90% to 100%). The conversion rate from RUS to PCN ranged from 10% to 42.6% (mean = 22.5%), while internalization of the PCN occurred in 11.7% to 98% of the patients (mean = 45.5%). Conclusions: Both diversional methods are effective in management of MUO. However, because of the heterogeneity of the included studies, the superiority of one of the procedures cannot be concluded.

Introduction

Although the true incidence is poorly defined [1], ureteral obstruction is a common manifestation of advanced urological and non-urological malignancies. It is a concerning sign and could be the first indication of cancer progression [2]. Malignant ureteral obstruction (MUO) can arise from direct invasion of the ureters by various abdominopelvic malignancies, encasement by malignant lymph nodes, or compression from an external malignant retroperitoneal mass [3,4,5]. This obstruction can appear as a dilatation in the ureter and or kidney. Urinary diversion (UD) is usually required for symptomatic relief and renal function preservation, especially when systemic therapy such as chemotherapy is planned [6,7].

While MUO is often detected incidentally at the time of abdominal imaging [8], symptoms at initial presentation can range from subacute nonspecific symptoms to acute pain, fever, vomiting, and sepsis [3,4]. A variety of imaging modalities are available to detect upper tract dilatation and determine the degree and the location of the ureteral obstruction. These include abdominal ultrasound, intravenous pyelography, abdominal CT, abdominal MRI, and retrograde pyelogram [9]. Abdominal CT with intravenous contrast is most effective in identifying the abdominal and pelvic pathologies responsible for the obstruction [8]. When both ureters are involved, prompt intervention may be required. If left untreated, bilateral obstruction can lead to uremia and electrolyte imbalance, with a potentially lethal outcome.

Selecting the optimal method of intervention to treat MUO is a clinical challenge for urologists and interventional radiologists. Many interventional procedures for decompression of the upper tract in patients with MUO have been described. However, insertion of percutaneous nephrostomy (PCN) and retrograde ureteral stent (RUS) are the most established procedures. These have been reported to have varying success, efficacy, and complication rates, and potentially differential impact on quality of life. Ureteral stenting is an attractive first-line option in principle because it is less invasive and the patient does not have to manage an external device; however, adequate drainage can be monitored more effectively with a PCN, and it is more easily changed. Differences in efficacy and quality of life associated with these procedures have not been clearly determined [10], and although they are performed in everyday practice, no guidelines have been established regarding the optimal method of UD [10,11,12]. The selection of one method over the other to decompress the upper tract in patients with MUO remains controversial. Here we review studies that compare PCN and RUS with respect to preservation of renal function and clinical outcomes.

Methods

All study types that compared PCN and RUS (head-to-head comparison) and contained original data were eligible for this review. Studies were included regardless of language and publication status, including abstracts without full text.

Studies were eligible if they included adult (≥ 18 years) subjects who underwent PCN or RUS to drain a kidney with MUO. Studies involving pediatric participants, pregnant women, animals, or only participants with benign ureteral obstruction (BUO) were excluded. Studies that assessed a mixed population of BUO and MUO were included.

Primary outcomes were preservation of the renal function, technical success rate, and complication rate. Secondary outcomes were conversion from one diversional method to another; residual hydronephrosis; length of hospital stay; procedure time; financial cost of the intervention; quality of life of patients post intervention.

Search method for identification of the studies

A comprehensive electronic search was performed in Ovid Medline, Ovid EMBASE, CINAHL, Cochrane Central Register of Controlled Trials (CENTRAL), and Scopus. Additionally, we looked for gray literature through OpenGray (https://opengrey.eu/), dissertation and thesis database (ProQuest) via (https://www.proquest.com). Also, we searched the Clinical trial.gov website and checked the reference lists of all the included studies. The date range for inclusion was inception of the database to October 2022. The result of the search was uploaded into Covidence software after potential duplicate studies were removed (identified using the reference management software RefWorks). The key words used in the search strategy were “neoplasm,” “hydronephrosis,” “percutaneous nephrostomy,” “ureteral stent.” Detailed search strategies are shown in the Online Appendix. The search result is summarized in a PRISMA flow chart (Figure 1).

Selection of the studies

Two authors (W.E. and M.R.) independently screened the title and the abstract of each study for its eligibility for the systematic review using Covidence software (Melbourne, Australia). Each study was marked “yes,” “no,” or “maybe” according to its relevance to the review. Any conflict in the screening was solved by a discussion between the 2 authors. A full text was obtained for all studies that were marked “yes” or “maybe” and independently reviewed by the 2 authors to decide which met the inclusion criteria. Google Translate was used to translate the full text of the studies that were published in languages other than English.

Data Extraction

The 2 authors extracted the data from the eligible studies into a data extraction sheet. These data include author names, year of publication, country of study, number of involved institutions, type of study, sample size, number of participants in each study arm, success rate, percentage and type of complications, mean and/or median age, creatinine at intervention and post intervention, residual hydronephrosis, percentage of conversion into another UD method, length of hospital stay, surgical time, cost of procedure in US dollars, and quality of life.

Quality assessment

The quality of the enrolled studies was assessed using the Newcastle-Ottawa Scale, which allocates a possible maximum of 9 stars for selection domain (4 items with 4 stars), comparability domain (1 item with 2 stars) and outcome domain (3 items with 3 stars). Studies with ≥ 7 stars are considered good quality, 6 stars fair quality, and ≤ 5 stars poor quality (Table 1).

Results and Discussion

Twenty-five studies were included: 21 reported on 1302 patients with MUO only [5,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32] (Table 1). Four of these 21 studies were published in a language other than English [17,24,25,28]. Three additional studies and one abstract reported on a total of 562 patients, of whom 388 had BUO and 174 had MUO [33,34,35,36] (Table 2). Because we could not get separate data for MUO patients (although the authors were contacted), the stated results in these 4 studies include both MUO and BUO patients. All but 2 of the included studies in the present review were conducted retrospectively [14,35]. The studies are summarized in Table 3.

Preservation of renal function

Few retrospective studies evaluated the difference in creatinine changes between the 2 groups. Ku et al. [20], Hyppolite et al. [23], Kurita et al. [28], and De Lorenzis et al. [32] investigated and reported the mean serum creatinine values before and after the intervention in patients treated with PCN and RUS (Table 4). Both UD methods successfully reduced creatinine and preserved renal function. Because of the greater reduction in creatinine after diversion by PCN compared with the RUS, Hyppolite et al. [23] stated that PCN was superior to RUS, especially in patients with bilateral PCN. The ratio of the creatinine reduction in the PCN was 78.8% compared with 70.7% in the RUS. This study has multiple limitations, including a small sample size, unequal groups, and lack of statistical analysis. Botkin et al. [31] supported this conclusion in their own study, in which the serum creatinine declined by 2.1 mg/dL in the PCN group and 0.3 mg/dL in the RUS group (PCN = 65.6%; RUS = 20%). However, the creatinine value before the intervention was at the upper limit of normal in the RUS group (1.5 mg/dL), but more than 2-fold higher in the PCN group (3.2 mg/dL). In contrast, the percentage of creatinine reduction was comparable between the 2 groups in the studies conducted by Kurita et al. [28] (PCN = 77.4%; RUS = 76.5%), Ku et al. [20] (PCN = 44.4%; RUS = 46.1%; P = 0.058), and De Lorenzis et al. [32] (PCN = 35.7%; RUS = 33.3%; P = 0.8). While no statistical analysis was performed by Kurita et al. [28], the statistically equivalent effect of PCN and RUS on the renal function was shown by Ku et al. [20] and De Lorenzis et al. [32]. Lack of reporting the time at which the creatinine was measured postoperatively might be the reason behind the variability in the percentage of the creatinine reduction among the studies. Among the above studies, only Kurita et al. [28] documented the time of the creatinine assessment, which was 2 weeks post intervention. Kanou et al. [18] noted that renal function improved after UD in all MUO patients, irrespective of the UD method.

Regarding the studies that assessed renal function in a mixed population (MUO and BUO), Ghous et al. [33] and Chang et al. [36] concluded that PCN is superior to RUS. This was based on the higher residual hydronephrosis in the RUS group compared with the PCN group.

Table 1. Characteristics of the included studies with MUO patients.

Table 1. Characteristics of the included studies with MUO patients.

Table 2. Characteristics of the enrolled studies that included both MUO and BUO patients.

Table 2. Characteristics of the enrolled studies that included both MUO and BUO patients.

Table 3. Summary of the primary tumors, objectives, and the conclusions of the included studies.

Table 3. Summary of the primary tumors, objectives, and the conclusions of the included studies.

Table 4. Serum creatinine changes.

Table 4. Serum creatinine changes.

Table 5. Technical success rate.

Table 5. Technical success rate.

Table 6. Complication rate.

Table 6. Complication rate.

Table 7. Conversion from one diversional method to the other.

Table 7. Conversion from one diversional method to the other.

Table 8. Length of hospital stay in days.

Table 8. Length of hospital stay in days.

Although RUS successfully improved the renal function in MUO, some studies reported that it failed to maintain this improvement, and PCN was required. Liaw et al. [22] revealed that PCN was necessary to maintain renal drainage in 38% of the RUS group. Tan et al. [13] and Song et al. [15] recorded 13% and 18% respectively of the RUS converted into PCN because they failed to drain the kidney. Ganatra et al. [11] also reported that 24% of 133 patients with RUS required PCN because of late failure of the stents as a result of cancer progression. Domico and Dewolf reported that 46% of patients with extrinsic ureteral obstruction failed to maintain kidney drainage within 30 days of insertion [37] Jenkins and Marcus reported that 2 of 10 stents lost their patency during an average observation period of 20 months in patients with MUO [38]. Chang et al. [36] evaluated the difference between RUS and PCN in maintaining renal function after diversion. They reported a lower elevation in creatinine in the PCN group (0.21 mg/dL) compared with the RUS group (0.78 mg/dL; P = 0.03) during the diversion period. Kanou et al. [18] found that using stents without shaft vent holes was successful in improving and maintaining renal function.

In summary, both diversional methods are effective, and there is inconclusive evidence to suggest the superiority of one intervention over the other. However, the available evidence suggests that PCN provides more durable renal drainage than RUS and therefore maintains renal function better.

Technical success rate

In this review, the technical success rate is defined as the successful insertion of a RUS or PCN. The success rate was reported in 14 studies in patients with MUO [5,13,15,16,18,19,21,22,24,26,27,29,30,32]. The success rate of RUS in these studies ranged from 21% to 100% (mean 70.3%). The success rate for PCN was higher in these studies, ranging from 90% to 100% (mean 98.8%). Additionally, one study [35] included patients with MUO or BUO. Here the success rate of RUS was 83% and of PCN was 92% (Table 5).

According to some studies, the primary tumor type significantly impacts the success rate of the RUS. RUS has a high failure rate in MUO caused by bladder and prostate cancer [6,8,19,30]. Wong et al. [19] and Zadra et al. [30] recommended that PCN should be the primary UD method in bladder and prostate cancers. However, Ganatra et al. [11] found that the primary tumor type did not determine the need for PCN. Song et al. [39] reached the same conclusion, stating that the type of gynecological tumor was not a predictor of the need for PCN, and he recommended that RUS should be the first option in MUO caused by gynecological tumors. Regarding the primary tumor in the other studies with a high RUS success rate, Liaw et al. [22] and Migita et al. [16] enrolled only gastric cancer patients in their studies, whereas Wong et al. [19] did not specify the initial tumor type of the included patients. While Kraemer et al. [17] did not report the success rate between the 2 groups in their cohort of 51 patients with prostate cancer, they reported that both UD methods were efficient in managing MUO in prostate cancer. Online Appendix Table 1 shows a summary of the studies based on the type of the primary tumor as well as the material, size, and exchange time of RUS and PCN.

On the other hand, a very low success rate has been recorded in other studies. Chitale et al. [21] found a success rate of 21% among 65 patients, of whom 30 had bladder cancer and 28 had prostate cancer. Gasparini et al. [29] reported a success rate of 39% in their cohort of 22 patients, of whom 6 had bladder cancer and 2 had prostate cancer. In the hands of experienced urologists, RUS in MUO can reach a high success rate. However, we recommend PCN as the preferable UD method in MUO caused by bladder cancer and, to a lesser degree, prostate cancer.

Complication rate

Different complications associated with RUS and PCN were reported in 15 publications out of 25 in this review (Table 6) [5,13,14,16,18,19,20,23,24,25,26,27,28,31,35]. Botkin et al. [31], Ku et al. [20], Song et al. [15], Tan et al. [13], Wong et al. [19], and Monsky et al. [14] reported no statistically significant difference in the complication rate between PCN and RUS groups. In RUS, pain and symptoms of bladder irritation were documented in 9 studies [5,14,16,18,25,26,28,31,35], ranging from 5% to 85% of the included patients (mean = 24%), whereas pain associated with PCN was reported in 4 studies from 1% to 31% (mean = 16%) [14,18,19,31]. Dislodgment of the PCN is a frequently encountered complication reported in 12 studies, varying between 3% and 44% (mean = 15.6%), [5,13,14,16,18,19,20,24,26,27,28,35], while migration of the ureteral stent was noted in 2% to 10% of the patients in 6 studies (mean = 6%) [5,13,14,26,27,35].

Infection associated with PCN and RUS was investigated in 10 series [14,16,19,20,23,25,27,28,31,35]. Ahmad et al. [35], Migita et al. [16], Hyppolite et al. [23], Stevens et al. [25], and Barton et al. [27] found infection to be more common with RUS, while Botkin et al. [31], Monsky et al. [14], Wong et al. [19], and Kurita et al. [28] reported that it was higher with PCN. Nonetheless, the difference between the 2 groups in the infection rate was not statistically significant. In the RUS group the infection ranged from 6% to 86% (mean = 25.5%), compared with 3.5% to 35% (mean = 16.7%) in the PCN group.

Stent obstruction and nephrostomy blockage were reported in 7 publications. This occurred from 11% to 54% (mean = 23%) in the RUS group, [16,18,19,20,24,28] and from 1% to 25% (mean = 14.9%) after PCN [14,18,19,20]. All these studies showed that obstruction was more common after RUS than PCN with the exception of Monsky et al. [14].

Six studies evaluated the difference in the rate of hematuria between PCN and RUS. Ahmad et al. [35] Migita et al. [16] Wong et al. [19], and Barton et al. [27] reported that hematuria is more common in the RUS groups, while Botkin et al. [31] and Kurita et al. [28] observed it more commonly in the PCN group. Bleeding from the PCN site was reported in 4% of cases by Kanou et al. [18] and 18% of the patients by Hyppolite et al. [23] reported on one patient with arterial hemorrhage requiring embolization after RUS insertion.

Other complications of RUS have been reported in a small number of studies. Stent breakage was found in 4.7% of cases by Stevens et al. [25] and 4% by Wong et al. [19]. Stone formation was reported in 2.7% by Kanou et al. [18]. Ureteral perforation was described in 9.5% of cases by Stevens et al. [25] and 1% by Ahmad et al. [35]. Fistula formation was documented in 4.7% (to the iliac artery) by Stevens et al. [25] and 6.5% (to the vagina) by Monsky et al. [14].

In the PCN group, fistula was reported in 3.5% by Desportes et al. [24]. Leakage around the PCN was seen in 2.5% to 18%: 2.5% by Wong et al. [19], 5% by Barton et al. [27] 6% by Monsky et al. [14] and 18% by Hyppolite et al. [23]. Urinary retention occurred equally frequently in both groups in one study: PCN 1.9% versus RUS 1.3% [31].

Some studies compared the percentage of complication between the 2 groups in general without detailing the type of complication [15,30]. All studies reported equivalent complication rates between the diversion types [13,16,19,20,31] except De Lorenzis et al. [32] who reported higher complications after RUS (PCN = 4.2%; RUS = 22.2%; P = 0.06). Overall, the global complication rate between procedures was comparable.

Conversion into another diversional method

While conversion from PCN to stent is the ultimate goal to eliminate external tubes and collecting devices, conversion from stent to PCN typically occurs secondary to stent complications or inability of the stent to drain the kidney. In this review, 8 studies showed switching from stents into PCNs, [5,13,15,18,22,31,33,36], and 7 studies showed internalization of the PCNs [19,21,24,25,26,29,31]. The rate of conversion into PCN ranged from 10% to 42.6% (mean = 22.5%), whereas internalization of the PCN ranged from 11.7% to 98% (mean = 45.5%) (Table 7).

Residual hydronephrosis

Only 2 studies explored the difference in the residual hydronephrosis between PCN and RUS [33,36]. These 2 studies included patients with MUO and BUO, and they revealed identical results, with 65.2% residual hydronephrosis in the RUS group and 27.2% in the PCN group.

Length of hospital stay

Five studies investigated the difference in hospital stay between PCN and RUS [13,15,17,24,32], and 3 of them showed a statistically significant duration of hospital stay in the patients with PCN [13,15,32]. This may be attributed to the decreased overall health of patients who undergo PCN (Table 8).

Procedure time

Three studies evaluated the difference in procedure time between the 2 groups. In all 3 studies, PCN required more time than RUS. Tan et al. [13], Song et al. [15], and Kanou et al. [18] reported mean procedure times of 30.6 ± 10.1, 39 ± 7.8, and 41.2 with RUS, and 51 ± 8.7, 52.4 ± 6.4, and 48.8 with PCN, respectively.

Financial cost

According to Tan et al. [13] and Song et al. [15], the PCN was more expensive than the RUS. The 2 studies were conducted in China in 2019 and 2012. The average cost was US$173.5 ± 4.1 and $89 ± 3.04 for RUS insertion compared with $595.14 ± 5.34 and $468 ± 3 for PCN insertion.

Quality of life assessment post intervention

Four studies compared the quality of life between PCN and RUS in patients with MUO [5,14,18,26]. However, only Monsky et al. [14] used a validated questionnaire to prospectively assess the quality of life (FACT-BL version 4). He assessed the emotional, functional, and physical well-being of patients after both types of UD at 7, 30, and 90 days post intervention and concluded that there was no statistically significant difference between the 2 groups. It is worth mentioning that the baseline quality of life was not assessed before intervention in both groups, which might induce bias in the result. “Useful life” was used to assess the improvement in the quality of life after the intervention in 2 studies [5,26]. It is calculated based on the presence of pain, complications, the ability to return home, and mental capacity. Feng et al. [5] reported that useful life was achieved in 82% of PCN patients and 87% of RUS patients. According to Hubner et al. [26] “useful life” was achieved in 96% of the RUS group and 68% of the PCN group. Kanou et al. [18] evaluated the difference in the quality of life between the 2 groups according to the number of days the patients spent out of the hospital. Only 60% of the RUS group patients left the hospital and were able to spend an average of 5.7 months out of it, whereas in the PCN group, 79% of the patients left and spent 4.5 months away from the hospital. Despite the heterogeneity in the assessment of the quality of life in MUO patients among the 4 included studies in this review, none of these studies showed a statistically significant difference in the quality of life between the 2 groups.

Limitations

The systematic review has several limitations, including an unequal and/or small sample size of the 2 groups in many of the included studies. In addition, the heterogeneity among the studies reduce the reliability of the evidence in the findings. Very few studies investigated and compared the changes in the renal function after intervention, and most of these were single-centre studies with significant limitations. The retrospective nature of the included studies and the absence of randomization increase the possibility of bias in these studies, especially with respect to selection bias in determining which patients receive RUS or PCN. Other limitations in some studies that limited their generalizability were the inclusion of a specific type of tumor to compare the PCN and RUS as in the studies that involved only gastric or gynecological cancers.

Conclusion

Both procedures are effective in management of MUO. Because of the heterogeneity of the included studies, the superiority of one of the procedures over the other cannot be concluded.

A future randomized control trial with a large sample size will be required to eliminate the bias of retrospective studies. In the absence of such a trial, a larger multicentre retrospective study with some form of propensity matching may provide a more accurate assessment of the differences. Special attention is required on the impact of these interventions on patient quality of life.