JOURNAL OF ENDOUROLOGY Volume 28, Number 6, June 2014 ª Mary Ann Liebert, Inc. Pp. 729–734 DOI: 10.1089/end.2013.0792
Metallic Ureteral Stents in Malignant Ureteral Obstruction: Clinical Factors Predicting Stent Failure Po-Ming Chow, MD,1 Jui-Shan Hsu, MD,2,3 Chao-Yuan Huang, MD, PhD,1 Shuo-Meng Wang, MD,1 Yuan-Ju Lee, MD,1 Kuo-How Huang, MD, PhD,1 Hong-Jheng Yu, MD, PhD,1 Yeong-Shiau Pu, MD, PhD,1 and Po-Chin, Liang, MD 2
Abstract
Purpose: To provide clinical outcomes of the Resonance metallic ureteral stent in patients with malignant ureteral obstruction, as well as clinical factors predicting stent failure. Methods: Cancer patients who have received Resonance stents from July 2009 to March 2012 for ureteral obstruction were included for chart review. Stent failure was detected by clinical symptoms, image studies, and renal function tests. Survival analysis for stent duration was used to estimate patency rate and factors predicting stent failure. Results: A total of 117 stents were inserted successfully into 94 ureteral units in 79 patients. There were no major complications. These stents underwent survival analysis and proportional hazard regression. The median duration for the stents was 5.77 months. In multivariate analysis, age (P = 0.043), preoperative serum creatinine level (P = 0.0174), and cancer type (P = 0.0494) were significant factors associated with stent failure. Cancer treatment before and after stent insertion had no effect on stent duration. Conclusions: Resonance stents are effective and safe in relieving malignant ureteral obstructions. Old age and high serum creatinine level are predictors for stent failure. Stents in patients with lower gastrointestinal cancers have longer functional duration. Introduction
U
reteral obstruction caused by extrinsic compression in malignant diseases has been challenging for urologists; various procedures have been used to improve urinary drainage. Relieving the obstruction with regular Double-J stents had high failure rates.1,2 Because the median life expectancy is generally less than 1 year in patients with metastatic cancer causing ureteral obstruction,3 new materials and techniques were introduced to maintain kidney function as well as quality of life. Segmental metal mesh stents seemed promising at the beginning,4 but long-term results were not satisfying.5,6 A unique design of ureteral stent is the Resonance metallic ureteral stent (Cook Medical, Bloomington, IN). Made of a continuous unfenestrated coil of nonmagnetic nickel-cobaltchromium-molybdenum alloy, the full-length stent allows a 1-year indwelling time, according to the manufacturer. The use of the Resonance stent was first reported in 2006 in a 64-year-old woman with metastatic breast cancer,7 followed by a series including both benign and malignant dis-
eases.8–18 Eight of these studies have included the follow-up results regarding malignant ureteral obstructions in up to 27 patients8–15; only three of them have analyzed risk factors of stent failure. We hereby provide our experience on stent duration with clinical factors predicting stent failure. Methods
The local Institutional Review Board approved this retrospective study and waived the informed consent requirement. In our hospital, the antegrade method of stent insertion is performed by radiologists under local anesthesia, and the retrograde method by urologists under general anesthesia. Routine follow-up studies include urinalysis, renal function test, plain radiography, and renal ultrasonography. Additional studies such as urine culture, antegrade pyelography, computed tomography (CT), or magnetic resonance imaging (MRI) are performed if clinically indicated. All cancer patients who have received Resonance stents from July 2009 to March 2012 for malignant ureteral obstruction were included for chart review. Patients with only
Departments of 1Urology and 2Medical Imaging, National Taiwan University Hospital and National Taiwan University College of Medicine, Taipei, Taiwan. 3 Department of Medical Imaging, Cardinal Tien Hospital, School of Medicine Fu-Jen Catholic University, New Taipei City, Taiwan.
729
730
failure attempts of insertion were not included. Bilateral ureters in the same patient were viewed as individual ureteral units (UUs), and repeated stent insertions into the same UU at different times were counted separately. Patient profiles and laboratory data were collected. Stent-related complications were reviewed. Urinary tract infection (UTI) was defined by a positive culture with urinary symptoms. Images before and after the procedure were reviewed to confirm the resolution and the recurrence of obstruction. The Kaplan-Meier model was used for survival analysis of the stents. The duration of each stent was counted from the day of insertion to the day of removal, or to the day when percutaneous nephrostomy (PCN) was performed. For those who decided not to have the stent removed because of terminal disease status, the date of stent failure was defined as the date of progressive hydronephrosis confirmed by image study. For patients who died or missed the follow-up before stent failure, the data were considered censored in survival analysis. Patient factors regarded as possible risk factors for stent failure include: Age (as continuous variable), age ‡ 60 years (as dichotomous variable), serum creatinine level (as continuous variable), serum creatinine ‡ 2 mg/dL (as dichotomous variable), cancer types, and treatment for cancer before and after stent insertion (operation, radiotherapy, chemotherapy, and target therapy). These were included for univariate and multivariate analysis using the Cox proportional hazard regression model. Statistical analyses were performed using free software (R version 2.14.1). All statistical tests were two-tailed, with P < 0.05 considered to indicate significance. To account for potential correlation between multiple stents in the same patient, we used methods that account for correlated data when plotting the Kaplan-Meier curve with robust variance estimates. Also, P-values from Cox models were based on robust variance, which accounts for correlated data. Results
From July 2009 to March 2012, 79 cancer patients with malignant ureteral obstruction underwent Resonance stent insertion in our hospital. These patients were followed until the end of May 2012. The male-to-female ratio was 22:57. The mean – standard deviation (SD) age was 59.9 – 12.2 years (range 32–85 years). The mean – SD body height was 157.7 – 7.3 cm (range 130–174.2 cm), and the mean – SD body weight was 55.2 – 13.7 kg (range 35.8–108.2 kg). The mean – SD body mass index body mass index was 22.1 – 4.76 kg/m2 (range 14.5–35.7 kg/m2). Cervical cancer was the most frequent type of malignancy (19 patients), followed by colorectal cancer (17 patients), gastric cancer (9 patients), and ovarian cancer (9 patients) (Table 1). A total of 117 stents were inserted successfully into 94 ureteral units in 79 patients. There were 74 stent failures (8 from UTI, 3 because of severe pain, 3 from migrations, and 60 because of progressive hydronephrosis) during follow-up. Of the management of these failures, 38 stents were replaced, 17 ureters underwent pigtail external drainage, 1 ureter received an ipsilateral stent, and 18 ureters had conservative treatment. The median functional duration was 5.77 months (95% confidence interval [CI] 4.03–8.7 months) (Fig. 1). The 3-month and 6-month patency rates were 65.16 – 5.71% and 44.4 – 6.69% (Table 2).
CHOW ET AL.
Table 1. Number of Patients, Ureteral Units, and Stents According to Type of Malignancy
Total GYN Cervical cancer Endometrial cancer Ovary cancer UGI Gastric cancer Pancreatic cancer Cancer of small intestine LGI Colon cancer Rectal cancer Cancer of appendix GU Bladder cancer Prostate cancer Others Anal cancer Breast cancer Lung cancer Lymphoma Sarcoma
Patient
Ureteral unit
Stent
79 29 19 1 9 12 9 2 1 18 11 6 1 10 5 5 10 2 2 1 4 1
94 34 23 1 10 15 11 2 2 21 13 7 1 11 6 5 13 2 3 1 5 2
117 44 31 2 11 17 12 2 3 31 20 9 2 11 6 5 14 2 3 2 5 2
GYN = gynecologic; UGI = upper gastrointestinal; LGI = lower gastrointestinal; GU = genitourinary.
Possible risk factors for stent failure were analyzed with Cox regression model. Age and serum creatinine level were both significant in univariate analysis. The hazard ratio was 1.938 (95% CI 1.079–3.482) for age ‡ 60 years and 2.599 (95% CI 1.42–4.757) for serum creatinine ‡ 2 mg/dL, respectively. Sex, degree of hydronephrosis, treatment for cancer, and occurrence of UTI were not significant (Table 3). After controlling for age and preoperative creatinine level, lower gastrointestinal (GI) cancer showed a lower risk than genitourinary (GU) cancer (P = 0.0494) (Table 4). The Kaplan-Meier curves showed differences of the duration of
FIG. 1. Kaplan-Meier curve for all stents (with 95% confidence interval).
6.43 (5–9.57) 71.15 – 5.60% 50.93 – 7.12% 4–476
Univariate analysis Hazard Lower Upper ratio 95% CI 95% CI
2.57 (1.33–5.37) 38.95 – 12.2% 16.69 – 8.28% 3–305 8.63 (5.37–11.07) 73.94 – 7.30% 58.74 – 9.40% 3–476 4.47 (2.5–5.77) 57.11 – 7.68% 31.35 – 7.54% 4–420
Age (continuous) Age ‡ 60 (dichotomous) Sex (male) Preoperative Cre (continuous) Preoperative Cre ‡ 2 (dichotomous) Cancer type GYN UGI LGI GU Degree of hydronephrosis Previous treatment Any Operation Radiotherapy Chemotherapy Target therapy Ongoing treatment Any Operation Radiotherapy Chemotherapy Target therapy UTI
Pvalue
1.03 1.938
1.004 1.079
1.056 3.482
0.0214* 0.0268*
1.103 1.181
0.5774 2.105 1.016 1.373
0.767 0.0303*
2.599
1.42
4.757
0.00195**
0.8653 1.946 0.5218 2.463 1.05
0.4748 0.8208 0.2815 1.0264 0.7479
1.577 4.614 0.9671 5.9101 1.475
0.636 0.131 0.0389* 0.0435* 0.777
0.4473 0.8535 0.9299 0.6378 0.9301
0.1574 0.4134 0.5211 0.3299 0.342
1.271 1.762 1.66 1.233 2.53
0.131 0.669 0.806 0.181 0.887
0.7173 1.532 1.272 0.6449 0.857 1.384
1.394 0.4541 0.6761 0.3576 0.3892 0.7723
0.3907 5.169 2.394 1.163 1.887 2.481
1.317 0.492 0.455 0.145 0.702 0.275
*p-value < 0.05; **p-value < 0.01. CI = confidence interval; Cre = creatinine; GYN = gynecologic; UGI = upper gastrointestinal; LGI = lower gastrointestinal; GU = genitourinary; UTI = urinary tract infection. CI = confidence interval; LGI = lower gastrointestinal; Cre = creatinine.
8.77 (5.8–11.1) 81.19 – 7.65% 68.23 – 9.66% 12–476 5.77 (4.03–8.7) 65.16 – 5.71% 44.4 – 6.69% 4–476 Median duration (95% CI) (months) 3-month patency 6-month patency Follow-up time (days)
731
Table 3. Univariate Analysis of Factors on Functional Duration
4.5 (2.93–6.23) 58.30 – 7.04% 32.95 – 7.75% 4–420
Age < 60 Age ‡ 60 Non-LGI LGI All
Table 2. Median Duration and Patency Rates of Stents in Different Subgroups
Cre ‡ 2
Cre < 2
UPPER URINARY TRACT OBSTRUCTION
stents according to cancer types, age, and preoperative creatinine level (Figs. 2–4). The median functional duration and patency rates of different groups were shown in Table 2. No major complications were reported. From the 117 stents, there were 10 patients with gross hematuria, 8 with dysuria, 3 with flank pain, and 5 with urinary frequency. UTI developed in 25 patients in 33 stents. There were three stent migrations in two patients. Discussion
In our study of 117 Resonance stents in 79 patients with malignant ureteral obstruction, the overall median functional duration was 5.77 months. Age, preoperative creatinine level, and cancer type were independent risk factors for stent failure in multivariate analysis. Sex, degree of hydronephrosis, cancer treatment before and after stent insertion, and occurrence of UTI were not associated with functional duration of the stents. The Resonance stent, with its unique design, is expected to offer longer durations against extrinsic compressions. The first report of successful clinical use of the Resonance stent was in a patient with retroperitoneal fibrosis resulting from
732
CHOW ET AL.
Table 4. Multivariate Analysis of Factors on Functional Duration Multivariate analysis Hazard ratio
Lower 95% CI
Upper 95% CI
Pvalue
Age 1.0285 1.0009 1.0568 0.043* Preoperative 1.2095 1.034 1.4147 0.0174* creatinine Cancer type GYN 0.6024 0.231 1.5709 0.3 UGI 1.0475 0.3767 2.9134 0.9291 LGI 0.3651 0.1337 0.9972 0.0494* GU Reference Reference Reference Reference Others 0.7912 0.295 2.1218 0.6417 *p-value < 0.05. CI = confidence interval; GYN = gynecologic; UGI = upper gastrointestinal; LGI = lower gastrointestinal; GU = genitourinary.
metastatic breast cancer, and the stent lasted 4 months.7 In most series, the median duration ranged from 3.5 to 11 months.9–13 In our study, the median duration for the Resonance stent was 5.77 months. Including the present study, no major complications have been reported. Although the results vary between the studies because of obvious differences in the patients’ backgrounds, sample size, and etiology of malignancy, the safety and efficacy of the Resonance stent is established. Before our study, three studies had assessed the risk factors associated with the failure of Resonance stents. In a cohort of 19 patients including malignant and benign obstructions, Wang and associates12 indicated that the proportion of patients with previous radiation therapy was significantly higher in the failure group. On the contrary, by analyzing 25 patients with malignant ureteral obstruction, Goldsmith and colleagues13 found that radiation was not associated with stent failure. They also found that GU cancer was a predictor
FIG. 2. Kaplan-Meier curve of stents grouped by cancer types. GU = genitourinary; GYN = gynecologic; LGI = lower gastrointestinal; UGI = upper gastrointestinal.
FIG. 3.
Kaplan-Meier curve of stents grouped by age.
of stent failure, but age was not in univariate analysis.13 Brown and coworkers14 observed more UTI in the obstruction group than in the patent group. In our study, age was an independent risk factor for stent failure in multivariate analysis, and radiation before or after stent insertion had no effect on stent duration; neither did UTI. Stents in patients with GU cancer had shorter duration compared with other types of cancers, and after further dividing cancers into different groups, stents in patients with lower GI cancer had significantly longer duration. The effect of preoperative creatinine level was not assessed in the previous studies and was shown to be related to stent failure for the first time. From our results, age, preoperative creatinine level, and cancer type were associated with stent duration. The effect of age on stent duration can be multiaspect. In cancer patients, older age might be related to more advanced disease and more rapid cancer progression, thus affecting stent durations. Patients with older age might also have more comorbidities and be more susceptible to UTI, resulting in more prompt stent replacement or external drainage than younger patients.
FIG. 4. Kaplan-Meier curve of stents grouped by preoperative creatinine level.
UPPER URINARY TRACT OBSTRUCTION
The relation between high preoperative serum creatinine level and stent failure might be inadequate urine production, which could cause encrustation on the stents more easily and thus affect stent duration. Liatsikos and colleagues10 observed encrustation in 12 of 54 stents in patients with benign and malignant disease, but the effect of encrustation on stent duration was not analyzed. Because of insufficient information about stent encrustation from medical records, we could not analyze it either. In our cohort, patients with lower GI malignancies had significantly longer stent durations, while treatment modality did not appear to be related. One possibility is that lower GI malignancies, either primary or metastatic, have texture and composition softer than other types of malignancies and cause less compression to ureters. Therefore, the Resonance stents could be more sustainable in patients with lower GI malignancies. These possible explanations remain only hypotheses, and further study is needed to investigate the association between these factors and stent duration. No studies yet have compared directly different metallic ureteral stents. One of the widely used metallic ureteral stents is the Memokath stent (PNN Medical, Denmark). The dual expansion stent is a semi-permanent, nickel-titanium alloy spiral stent of varying lengths (120–200 mm) with a lumen.19 In a series that consisted of 73 patients, the average indwelling time of an individual Memokath was 11.2 months.20 One major structural difference between the Memokath stent and the Resonance stent is that the Memokath stent has a lumen, but the Resonance stent does not. Further study is needed to investigate the advantages of metallic stents with or without lumens. Nonetheless, the unique design of the Memokath 051 stent produced significantly less stent-related urinary symptoms compared with conventional Double-J–shaped stents.21 The definition and detection of stent failure is largely clinical: We take both imaging and functional studies into account when deciding to remove a stent or place a PCN. For those who need hospice care and refuse further diversion (which consisted of 18 stents in 14 patients in our cohort), the date of progressive hydronephrosis was defined as the date of stent failure. The progressive hydronephrosis was often detected earlier than marked elevation of serum creatinine. One explanation could be the compensation of the contralateral functioning kidney, stabilizing overall renal function after stent failure in one side. Another explanation could be the intentionally infrequent blood test in those patients, resulting in a delayed confirmation of creatinine elevation. In any case, stable residual hydronephrosis would not be regarded as a sign of stent failure. To confirm the occurrence of stent dysfunction is not always easy in clinical practice and to accurately define the specific time point of stent failure is even more challenging. In our retrospective study, the date of stent removal or PCN insertion was used as an approximation or surrogate of the actual failure date, and this is also the definition used by previous series. Using the same criteria is helpful for comparison between different studies. To better define stent failure, however, a strict follow-up protocol in a prospective cohort is needed. The decision whether to receive a Resonance stent should be made between the patient and the clinician. The aim of our study is to provide more information regarding the functional outcome of the stent and possible risks of failure. Whether
733
these outcomes meet the expectation of the patient could vary between individuals. Patients older than 60 years (median patency 4.47 months) might still benefit more from the Resonance stent than regular stents or external drainage, while the benefit might be less significant in patients with a preoperative serum creatinine level > 2 mg/dL (median patency 2.57 months). Permanent PCN seems to affect quality of life more; therefore, it is often reserved as a second-line therapy. Compared with the other studies, there are several advantages of our study: A relatively large patient number receiving Resonance stents, a relatively homogenous cohort consisting of only cancer patients, the use of survival analysis for duration of the stents instead of crude failure rate, and the use of a multivariate regression model to control confounding factors. Our study does have some limitations, however. First, this is a retrospective study. Second, instead of receiving Resonance stents as soon as hydronephrosis developed, most patients had failure of regular polymeric stents before they had metallic stents. This delayed use of a metallic stent might make the duration of the stents shorter than it should have been; the cancer might have progressed, so there is a potential selection bias from the cohort. Third, there were three migrated stents that could have stayed longer in the ureter if they had not migrated. Whether a migrated stent should be considered a failure remains controversial, and the functional duration would be slightly increased if it were regarded as functioning. Finally, the number of cases was too small to analyze all possible factors. Future studies can adopt a prospective protocol to find more factors that might influence stent duration, and these factors can be examined in patients receiving other types of stents. Conclusion
The Resonance ureteral stent is safe and effective in relieving malignant ureteral obstructions. Age and preoperative serum creatinine level were independent risk factors for stent failure. Lower GI cancers had longer stent duration than GU cancers. The benefits of Resonance stents can be maximized under careful patient selection. Acknowledgments
We would like to express our appreciation to Chia-Yen Chen from Harvard School of Public Health for his help in statistical analysis and Joanna Yeh from Sphera International for language supports. Disclosure Statement
No competing financial interests exist. References
1. Docimo SG, Dewolf WC. High failure rate of indwelling ureteral stents in patients with extrinsic obstruction: Experience at 2 institutions. J Urol 1989;142:277–279. 2. Chung SY, Stein RJ, Landsittel D, et al. 15-year experience with the management of extrinsic ureteral obstruction with indwelling ureteral stents. J Urol 2004;172:592–595. 3. Russo P. Urologic emergencies in the cancer patient. Semin Oncol 2000;27:284–298. 4. Lugmayr H, Pauer W. Self-expanding metal stents for palliative treatment of malignant ureteral obstruction. AJR Am J Roentgenol 1992;159:1091–1094.
734
5. Liatsikos EN, Karnabatidis D, Katsanos K, et al. Ureteral metal stents: 10-year experience with malignant ureteral obstruction treatment. J Urol 2009;182:2613–2617. 6. Sountoulides P, Kaplan A, Kaufmann OG, Sofikitis N. Current status of metal stents for managing malignant ureteric obstruction. BJU Int 2010;105:1066–1072. 7. Borin JF, Melamud O, Clayman RV. Initial experience with full-length metal stent to relieve malignant ureteral obstruction. J Endourol 2006;20:300–304. 8. Wah TM, Irving HC, Cartledge J. Initial experience with the resonance metallic stent for antegrade ureteric stenting. Cardiovasc Intervent Radiol 2007;30:705–710. 9. Nagele U, Kuczyk MA, Horstmann M, et al. Initial clinical experience with full-length metal ureteral stents for obstructive ureteral stenosis. World J Urol 2008;26:257–262. 10. Liatsikos E, Kallidonis P, Kyriazis I, et al. Ureteral obstruction: Is the full metallic double-pigtail stent the way to go? Eur Urol 2010;57:480–486. 11. Benson AD, Taylor ER, Schwartz BF. Metal ureteral stent for benign and malignant ureteral obstruction. J Urol 2011;185: 2217–2222. 12. Wang HJ, Lee TY, Luo HL, et al. Application of resonance metallic stents for ureteral obstruction. BJU Int 2011;108: 428–432. 13. Goldsmith ZG, Wang AJ, Banez LL, et al. Outcomes of metallic stents for malignant ureteral obstruction. J Urol 2012;188:851–855. 14. Brown JA, Powell CL, Carlson KR. Metallic full-length ureteral stents: Does urinary tract infection cause obstruction? ScientificWorldJournal 2010;10:1566–1573. 15. Kadlec AO, Ellimoottil CS, Greco KA, Turk TM. Five-year experience with metallic stents for chronic ureteral obstruction. J Urol 2013;190:937–941. 16. Garg T, Guralnick ML, Langenstroer P, et al. Resonance metallic ureteral stents do not successfully treat ureteroenteric strictures. J Endourol 2009;23:1199–1202. 17. Lopez-Huertas HL, Polcari AJ, Acosta-Miranda A, Turk TM. Metallic ureteral stents: A cost-effective method of managing benign upper tract obstruction. J Endourol 2010; 24:483–485.
CHOW ET AL.
18. Modi AP, Ritch CR, Arend D, et al. Multicenter experience with metallic ureteral stents for malignant and chronic benign ureteral obstruction. J Endourol 2010;24:1189–1193. 19. Masood J, Papatsoris A, Buchholz N. Dual expansion nickel-titanium alloy metal ureteric stent: Novel use of a metallic stent to bridge the ureter in the minimally invasive management of complex ureteric and pelviureteric junction strictures. Urol Int 2010;84:477–478. 20. Papatsoris AG, Buchholz N. A novel thermo-expandable ureteral metal stent for the minimally invasive management of ureteral strictures. J Endourol 2010;24:487–491. 21. Maan Z, Patel D, Moraitis K, et al. Comparison of stentrelated symptoms between conventional Double-J stents and a new-generation thermoexpandable segmental metallic stent: A validated-questionnaire-based study. J Endourol 2010;24:589–593.
Address correspondence to: Po-Chin Liang, MD Department of Medical Imaging National Taiwan University Hospital No. 1, Changde Street Zhongzheng District Taipei City 10048 Taiwan E-mail: [email protected] Abbreviations Used CI ¼ confidence interval CT ¼ computed tomography GI ¼ gastrointestinal GU ¼ genitourinary MRI ¼ magnetic resonance imaging PCN ¼ percutaneous nephrostomy SD ¼ standard deviation UTI ¼ urinary tract infection UU ¼ ureteral unit
This article has been cited by: 1. Ming-Hong Kao, Chung-Cheng Wang. 2014. The efficacy and safety of ureteral dilation and long-term type ureteral stent for patients with ureteral obstruction. Urological Science . [CrossRef]
Metallic Ureteral Stents in Malignant Ureteral Obstruction: Clinical Factors Predicting Stent Failure Po-Ming Chow, MD,1 Jui-Shan Hsu, MD,2,3 Chao-Yuan Huang, MD, PhD,1 Shuo-Meng Wang, MD,1 Yuan-Ju Lee, MD,1 Kuo-How Huang, MD, PhD,1 Hong-Jheng Yu, MD, PhD,1 Yeong-Shiau Pu, MD, PhD,1 and Po-Chin, Liang, MD 2
Abstract
Purpose: To provide clinical outcomes of the Resonance metallic ureteral stent in patients with malignant ureteral obstruction, as well as clinical factors predicting stent failure. Methods: Cancer patients who have received Resonance stents from July 2009 to March 2012 for ureteral obstruction were included for chart review. Stent failure was detected by clinical symptoms, image studies, and renal function tests. Survival analysis for stent duration was used to estimate patency rate and factors predicting stent failure. Results: A total of 117 stents were inserted successfully into 94 ureteral units in 79 patients. There were no major complications. These stents underwent survival analysis and proportional hazard regression. The median duration for the stents was 5.77 months. In multivariate analysis, age (P = 0.043), preoperative serum creatinine level (P = 0.0174), and cancer type (P = 0.0494) were significant factors associated with stent failure. Cancer treatment before and after stent insertion had no effect on stent duration. Conclusions: Resonance stents are effective and safe in relieving malignant ureteral obstructions. Old age and high serum creatinine level are predictors for stent failure. Stents in patients with lower gastrointestinal cancers have longer functional duration. Introduction
U
reteral obstruction caused by extrinsic compression in malignant diseases has been challenging for urologists; various procedures have been used to improve urinary drainage. Relieving the obstruction with regular Double-J stents had high failure rates.1,2 Because the median life expectancy is generally less than 1 year in patients with metastatic cancer causing ureteral obstruction,3 new materials and techniques were introduced to maintain kidney function as well as quality of life. Segmental metal mesh stents seemed promising at the beginning,4 but long-term results were not satisfying.5,6 A unique design of ureteral stent is the Resonance metallic ureteral stent (Cook Medical, Bloomington, IN). Made of a continuous unfenestrated coil of nonmagnetic nickel-cobaltchromium-molybdenum alloy, the full-length stent allows a 1-year indwelling time, according to the manufacturer. The use of the Resonance stent was first reported in 2006 in a 64-year-old woman with metastatic breast cancer,7 followed by a series including both benign and malignant dis-
eases.8–18 Eight of these studies have included the follow-up results regarding malignant ureteral obstructions in up to 27 patients8–15; only three of them have analyzed risk factors of stent failure. We hereby provide our experience on stent duration with clinical factors predicting stent failure. Methods
The local Institutional Review Board approved this retrospective study and waived the informed consent requirement. In our hospital, the antegrade method of stent insertion is performed by radiologists under local anesthesia, and the retrograde method by urologists under general anesthesia. Routine follow-up studies include urinalysis, renal function test, plain radiography, and renal ultrasonography. Additional studies such as urine culture, antegrade pyelography, computed tomography (CT), or magnetic resonance imaging (MRI) are performed if clinically indicated. All cancer patients who have received Resonance stents from July 2009 to March 2012 for malignant ureteral obstruction were included for chart review. Patients with only
Departments of 1Urology and 2Medical Imaging, National Taiwan University Hospital and National Taiwan University College of Medicine, Taipei, Taiwan. 3 Department of Medical Imaging, Cardinal Tien Hospital, School of Medicine Fu-Jen Catholic University, New Taipei City, Taiwan.
729
730
failure attempts of insertion were not included. Bilateral ureters in the same patient were viewed as individual ureteral units (UUs), and repeated stent insertions into the same UU at different times were counted separately. Patient profiles and laboratory data were collected. Stent-related complications were reviewed. Urinary tract infection (UTI) was defined by a positive culture with urinary symptoms. Images before and after the procedure were reviewed to confirm the resolution and the recurrence of obstruction. The Kaplan-Meier model was used for survival analysis of the stents. The duration of each stent was counted from the day of insertion to the day of removal, or to the day when percutaneous nephrostomy (PCN) was performed. For those who decided not to have the stent removed because of terminal disease status, the date of stent failure was defined as the date of progressive hydronephrosis confirmed by image study. For patients who died or missed the follow-up before stent failure, the data were considered censored in survival analysis. Patient factors regarded as possible risk factors for stent failure include: Age (as continuous variable), age ‡ 60 years (as dichotomous variable), serum creatinine level (as continuous variable), serum creatinine ‡ 2 mg/dL (as dichotomous variable), cancer types, and treatment for cancer before and after stent insertion (operation, radiotherapy, chemotherapy, and target therapy). These were included for univariate and multivariate analysis using the Cox proportional hazard regression model. Statistical analyses were performed using free software (R version 2.14.1). All statistical tests were two-tailed, with P < 0.05 considered to indicate significance. To account for potential correlation between multiple stents in the same patient, we used methods that account for correlated data when plotting the Kaplan-Meier curve with robust variance estimates. Also, P-values from Cox models were based on robust variance, which accounts for correlated data. Results
From July 2009 to March 2012, 79 cancer patients with malignant ureteral obstruction underwent Resonance stent insertion in our hospital. These patients were followed until the end of May 2012. The male-to-female ratio was 22:57. The mean – standard deviation (SD) age was 59.9 – 12.2 years (range 32–85 years). The mean – SD body height was 157.7 – 7.3 cm (range 130–174.2 cm), and the mean – SD body weight was 55.2 – 13.7 kg (range 35.8–108.2 kg). The mean – SD body mass index body mass index was 22.1 – 4.76 kg/m2 (range 14.5–35.7 kg/m2). Cervical cancer was the most frequent type of malignancy (19 patients), followed by colorectal cancer (17 patients), gastric cancer (9 patients), and ovarian cancer (9 patients) (Table 1). A total of 117 stents were inserted successfully into 94 ureteral units in 79 patients. There were 74 stent failures (8 from UTI, 3 because of severe pain, 3 from migrations, and 60 because of progressive hydronephrosis) during follow-up. Of the management of these failures, 38 stents were replaced, 17 ureters underwent pigtail external drainage, 1 ureter received an ipsilateral stent, and 18 ureters had conservative treatment. The median functional duration was 5.77 months (95% confidence interval [CI] 4.03–8.7 months) (Fig. 1). The 3-month and 6-month patency rates were 65.16 – 5.71% and 44.4 – 6.69% (Table 2).
CHOW ET AL.
Table 1. Number of Patients, Ureteral Units, and Stents According to Type of Malignancy
Total GYN Cervical cancer Endometrial cancer Ovary cancer UGI Gastric cancer Pancreatic cancer Cancer of small intestine LGI Colon cancer Rectal cancer Cancer of appendix GU Bladder cancer Prostate cancer Others Anal cancer Breast cancer Lung cancer Lymphoma Sarcoma
Patient
Ureteral unit
Stent
79 29 19 1 9 12 9 2 1 18 11 6 1 10 5 5 10 2 2 1 4 1
94 34 23 1 10 15 11 2 2 21 13 7 1 11 6 5 13 2 3 1 5 2
117 44 31 2 11 17 12 2 3 31 20 9 2 11 6 5 14 2 3 2 5 2
GYN = gynecologic; UGI = upper gastrointestinal; LGI = lower gastrointestinal; GU = genitourinary.
Possible risk factors for stent failure were analyzed with Cox regression model. Age and serum creatinine level were both significant in univariate analysis. The hazard ratio was 1.938 (95% CI 1.079–3.482) for age ‡ 60 years and 2.599 (95% CI 1.42–4.757) for serum creatinine ‡ 2 mg/dL, respectively. Sex, degree of hydronephrosis, treatment for cancer, and occurrence of UTI were not significant (Table 3). After controlling for age and preoperative creatinine level, lower gastrointestinal (GI) cancer showed a lower risk than genitourinary (GU) cancer (P = 0.0494) (Table 4). The Kaplan-Meier curves showed differences of the duration of
FIG. 1. Kaplan-Meier curve for all stents (with 95% confidence interval).
6.43 (5–9.57) 71.15 – 5.60% 50.93 – 7.12% 4–476
Univariate analysis Hazard Lower Upper ratio 95% CI 95% CI
2.57 (1.33–5.37) 38.95 – 12.2% 16.69 – 8.28% 3–305 8.63 (5.37–11.07) 73.94 – 7.30% 58.74 – 9.40% 3–476 4.47 (2.5–5.77) 57.11 – 7.68% 31.35 – 7.54% 4–420
Age (continuous) Age ‡ 60 (dichotomous) Sex (male) Preoperative Cre (continuous) Preoperative Cre ‡ 2 (dichotomous) Cancer type GYN UGI LGI GU Degree of hydronephrosis Previous treatment Any Operation Radiotherapy Chemotherapy Target therapy Ongoing treatment Any Operation Radiotherapy Chemotherapy Target therapy UTI
Pvalue
1.03 1.938
1.004 1.079
1.056 3.482
0.0214* 0.0268*
1.103 1.181
0.5774 2.105 1.016 1.373
0.767 0.0303*
2.599
1.42
4.757
0.00195**
0.8653 1.946 0.5218 2.463 1.05
0.4748 0.8208 0.2815 1.0264 0.7479
1.577 4.614 0.9671 5.9101 1.475
0.636 0.131 0.0389* 0.0435* 0.777
0.4473 0.8535 0.9299 0.6378 0.9301
0.1574 0.4134 0.5211 0.3299 0.342
1.271 1.762 1.66 1.233 2.53
0.131 0.669 0.806 0.181 0.887
0.7173 1.532 1.272 0.6449 0.857 1.384
1.394 0.4541 0.6761 0.3576 0.3892 0.7723
0.3907 5.169 2.394 1.163 1.887 2.481
1.317 0.492 0.455 0.145 0.702 0.275
*p-value < 0.05; **p-value < 0.01. CI = confidence interval; Cre = creatinine; GYN = gynecologic; UGI = upper gastrointestinal; LGI = lower gastrointestinal; GU = genitourinary; UTI = urinary tract infection. CI = confidence interval; LGI = lower gastrointestinal; Cre = creatinine.
8.77 (5.8–11.1) 81.19 – 7.65% 68.23 – 9.66% 12–476 5.77 (4.03–8.7) 65.16 – 5.71% 44.4 – 6.69% 4–476 Median duration (95% CI) (months) 3-month patency 6-month patency Follow-up time (days)
731
Table 3. Univariate Analysis of Factors on Functional Duration
4.5 (2.93–6.23) 58.30 – 7.04% 32.95 – 7.75% 4–420
Age < 60 Age ‡ 60 Non-LGI LGI All
Table 2. Median Duration and Patency Rates of Stents in Different Subgroups
Cre ‡ 2
Cre < 2
UPPER URINARY TRACT OBSTRUCTION
stents according to cancer types, age, and preoperative creatinine level (Figs. 2–4). The median functional duration and patency rates of different groups were shown in Table 2. No major complications were reported. From the 117 stents, there were 10 patients with gross hematuria, 8 with dysuria, 3 with flank pain, and 5 with urinary frequency. UTI developed in 25 patients in 33 stents. There were three stent migrations in two patients. Discussion
In our study of 117 Resonance stents in 79 patients with malignant ureteral obstruction, the overall median functional duration was 5.77 months. Age, preoperative creatinine level, and cancer type were independent risk factors for stent failure in multivariate analysis. Sex, degree of hydronephrosis, cancer treatment before and after stent insertion, and occurrence of UTI were not associated with functional duration of the stents. The Resonance stent, with its unique design, is expected to offer longer durations against extrinsic compressions. The first report of successful clinical use of the Resonance stent was in a patient with retroperitoneal fibrosis resulting from
732
CHOW ET AL.
Table 4. Multivariate Analysis of Factors on Functional Duration Multivariate analysis Hazard ratio
Lower 95% CI
Upper 95% CI
Pvalue
Age 1.0285 1.0009 1.0568 0.043* Preoperative 1.2095 1.034 1.4147 0.0174* creatinine Cancer type GYN 0.6024 0.231 1.5709 0.3 UGI 1.0475 0.3767 2.9134 0.9291 LGI 0.3651 0.1337 0.9972 0.0494* GU Reference Reference Reference Reference Others 0.7912 0.295 2.1218 0.6417 *p-value < 0.05. CI = confidence interval; GYN = gynecologic; UGI = upper gastrointestinal; LGI = lower gastrointestinal; GU = genitourinary.
metastatic breast cancer, and the stent lasted 4 months.7 In most series, the median duration ranged from 3.5 to 11 months.9–13 In our study, the median duration for the Resonance stent was 5.77 months. Including the present study, no major complications have been reported. Although the results vary between the studies because of obvious differences in the patients’ backgrounds, sample size, and etiology of malignancy, the safety and efficacy of the Resonance stent is established. Before our study, three studies had assessed the risk factors associated with the failure of Resonance stents. In a cohort of 19 patients including malignant and benign obstructions, Wang and associates12 indicated that the proportion of patients with previous radiation therapy was significantly higher in the failure group. On the contrary, by analyzing 25 patients with malignant ureteral obstruction, Goldsmith and colleagues13 found that radiation was not associated with stent failure. They also found that GU cancer was a predictor
FIG. 2. Kaplan-Meier curve of stents grouped by cancer types. GU = genitourinary; GYN = gynecologic; LGI = lower gastrointestinal; UGI = upper gastrointestinal.
FIG. 3.
Kaplan-Meier curve of stents grouped by age.
of stent failure, but age was not in univariate analysis.13 Brown and coworkers14 observed more UTI in the obstruction group than in the patent group. In our study, age was an independent risk factor for stent failure in multivariate analysis, and radiation before or after stent insertion had no effect on stent duration; neither did UTI. Stents in patients with GU cancer had shorter duration compared with other types of cancers, and after further dividing cancers into different groups, stents in patients with lower GI cancer had significantly longer duration. The effect of preoperative creatinine level was not assessed in the previous studies and was shown to be related to stent failure for the first time. From our results, age, preoperative creatinine level, and cancer type were associated with stent duration. The effect of age on stent duration can be multiaspect. In cancer patients, older age might be related to more advanced disease and more rapid cancer progression, thus affecting stent durations. Patients with older age might also have more comorbidities and be more susceptible to UTI, resulting in more prompt stent replacement or external drainage than younger patients.
FIG. 4. Kaplan-Meier curve of stents grouped by preoperative creatinine level.
UPPER URINARY TRACT OBSTRUCTION
The relation between high preoperative serum creatinine level and stent failure might be inadequate urine production, which could cause encrustation on the stents more easily and thus affect stent duration. Liatsikos and colleagues10 observed encrustation in 12 of 54 stents in patients with benign and malignant disease, but the effect of encrustation on stent duration was not analyzed. Because of insufficient information about stent encrustation from medical records, we could not analyze it either. In our cohort, patients with lower GI malignancies had significantly longer stent durations, while treatment modality did not appear to be related. One possibility is that lower GI malignancies, either primary or metastatic, have texture and composition softer than other types of malignancies and cause less compression to ureters. Therefore, the Resonance stents could be more sustainable in patients with lower GI malignancies. These possible explanations remain only hypotheses, and further study is needed to investigate the association between these factors and stent duration. No studies yet have compared directly different metallic ureteral stents. One of the widely used metallic ureteral stents is the Memokath stent (PNN Medical, Denmark). The dual expansion stent is a semi-permanent, nickel-titanium alloy spiral stent of varying lengths (120–200 mm) with a lumen.19 In a series that consisted of 73 patients, the average indwelling time of an individual Memokath was 11.2 months.20 One major structural difference between the Memokath stent and the Resonance stent is that the Memokath stent has a lumen, but the Resonance stent does not. Further study is needed to investigate the advantages of metallic stents with or without lumens. Nonetheless, the unique design of the Memokath 051 stent produced significantly less stent-related urinary symptoms compared with conventional Double-J–shaped stents.21 The definition and detection of stent failure is largely clinical: We take both imaging and functional studies into account when deciding to remove a stent or place a PCN. For those who need hospice care and refuse further diversion (which consisted of 18 stents in 14 patients in our cohort), the date of progressive hydronephrosis was defined as the date of stent failure. The progressive hydronephrosis was often detected earlier than marked elevation of serum creatinine. One explanation could be the compensation of the contralateral functioning kidney, stabilizing overall renal function after stent failure in one side. Another explanation could be the intentionally infrequent blood test in those patients, resulting in a delayed confirmation of creatinine elevation. In any case, stable residual hydronephrosis would not be regarded as a sign of stent failure. To confirm the occurrence of stent dysfunction is not always easy in clinical practice and to accurately define the specific time point of stent failure is even more challenging. In our retrospective study, the date of stent removal or PCN insertion was used as an approximation or surrogate of the actual failure date, and this is also the definition used by previous series. Using the same criteria is helpful for comparison between different studies. To better define stent failure, however, a strict follow-up protocol in a prospective cohort is needed. The decision whether to receive a Resonance stent should be made between the patient and the clinician. The aim of our study is to provide more information regarding the functional outcome of the stent and possible risks of failure. Whether
733
these outcomes meet the expectation of the patient could vary between individuals. Patients older than 60 years (median patency 4.47 months) might still benefit more from the Resonance stent than regular stents or external drainage, while the benefit might be less significant in patients with a preoperative serum creatinine level > 2 mg/dL (median patency 2.57 months). Permanent PCN seems to affect quality of life more; therefore, it is often reserved as a second-line therapy. Compared with the other studies, there are several advantages of our study: A relatively large patient number receiving Resonance stents, a relatively homogenous cohort consisting of only cancer patients, the use of survival analysis for duration of the stents instead of crude failure rate, and the use of a multivariate regression model to control confounding factors. Our study does have some limitations, however. First, this is a retrospective study. Second, instead of receiving Resonance stents as soon as hydronephrosis developed, most patients had failure of regular polymeric stents before they had metallic stents. This delayed use of a metallic stent might make the duration of the stents shorter than it should have been; the cancer might have progressed, so there is a potential selection bias from the cohort. Third, there were three migrated stents that could have stayed longer in the ureter if they had not migrated. Whether a migrated stent should be considered a failure remains controversial, and the functional duration would be slightly increased if it were regarded as functioning. Finally, the number of cases was too small to analyze all possible factors. Future studies can adopt a prospective protocol to find more factors that might influence stent duration, and these factors can be examined in patients receiving other types of stents. Conclusion
The Resonance ureteral stent is safe and effective in relieving malignant ureteral obstructions. Age and preoperative serum creatinine level were independent risk factors for stent failure. Lower GI cancers had longer stent duration than GU cancers. The benefits of Resonance stents can be maximized under careful patient selection. Acknowledgments
We would like to express our appreciation to Chia-Yen Chen from Harvard School of Public Health for his help in statistical analysis and Joanna Yeh from Sphera International for language supports. Disclosure Statement
No competing financial interests exist. References
1. Docimo SG, Dewolf WC. High failure rate of indwelling ureteral stents in patients with extrinsic obstruction: Experience at 2 institutions. J Urol 1989;142:277–279. 2. Chung SY, Stein RJ, Landsittel D, et al. 15-year experience with the management of extrinsic ureteral obstruction with indwelling ureteral stents. J Urol 2004;172:592–595. 3. Russo P. Urologic emergencies in the cancer patient. Semin Oncol 2000;27:284–298. 4. Lugmayr H, Pauer W. Self-expanding metal stents for palliative treatment of malignant ureteral obstruction. AJR Am J Roentgenol 1992;159:1091–1094.
734
5. Liatsikos EN, Karnabatidis D, Katsanos K, et al. Ureteral metal stents: 10-year experience with malignant ureteral obstruction treatment. J Urol 2009;182:2613–2617. 6. Sountoulides P, Kaplan A, Kaufmann OG, Sofikitis N. Current status of metal stents for managing malignant ureteric obstruction. BJU Int 2010;105:1066–1072. 7. Borin JF, Melamud O, Clayman RV. Initial experience with full-length metal stent to relieve malignant ureteral obstruction. J Endourol 2006;20:300–304. 8. Wah TM, Irving HC, Cartledge J. Initial experience with the resonance metallic stent for antegrade ureteric stenting. Cardiovasc Intervent Radiol 2007;30:705–710. 9. Nagele U, Kuczyk MA, Horstmann M, et al. Initial clinical experience with full-length metal ureteral stents for obstructive ureteral stenosis. World J Urol 2008;26:257–262. 10. Liatsikos E, Kallidonis P, Kyriazis I, et al. Ureteral obstruction: Is the full metallic double-pigtail stent the way to go? Eur Urol 2010;57:480–486. 11. Benson AD, Taylor ER, Schwartz BF. Metal ureteral stent for benign and malignant ureteral obstruction. J Urol 2011;185: 2217–2222. 12. Wang HJ, Lee TY, Luo HL, et al. Application of resonance metallic stents for ureteral obstruction. BJU Int 2011;108: 428–432. 13. Goldsmith ZG, Wang AJ, Banez LL, et al. Outcomes of metallic stents for malignant ureteral obstruction. J Urol 2012;188:851–855. 14. Brown JA, Powell CL, Carlson KR. Metallic full-length ureteral stents: Does urinary tract infection cause obstruction? ScientificWorldJournal 2010;10:1566–1573. 15. Kadlec AO, Ellimoottil CS, Greco KA, Turk TM. Five-year experience with metallic stents for chronic ureteral obstruction. J Urol 2013;190:937–941. 16. Garg T, Guralnick ML, Langenstroer P, et al. Resonance metallic ureteral stents do not successfully treat ureteroenteric strictures. J Endourol 2009;23:1199–1202. 17. Lopez-Huertas HL, Polcari AJ, Acosta-Miranda A, Turk TM. Metallic ureteral stents: A cost-effective method of managing benign upper tract obstruction. J Endourol 2010; 24:483–485.
CHOW ET AL.
18. Modi AP, Ritch CR, Arend D, et al. Multicenter experience with metallic ureteral stents for malignant and chronic benign ureteral obstruction. J Endourol 2010;24:1189–1193. 19. Masood J, Papatsoris A, Buchholz N. Dual expansion nickel-titanium alloy metal ureteric stent: Novel use of a metallic stent to bridge the ureter in the minimally invasive management of complex ureteric and pelviureteric junction strictures. Urol Int 2010;84:477–478. 20. Papatsoris AG, Buchholz N. A novel thermo-expandable ureteral metal stent for the minimally invasive management of ureteral strictures. J Endourol 2010;24:487–491. 21. Maan Z, Patel D, Moraitis K, et al. Comparison of stentrelated symptoms between conventional Double-J stents and a new-generation thermoexpandable segmental metallic stent: A validated-questionnaire-based study. J Endourol 2010;24:589–593.
Address correspondence to: Po-Chin Liang, MD Department of Medical Imaging National Taiwan University Hospital No. 1, Changde Street Zhongzheng District Taipei City 10048 Taiwan E-mail: [email protected] Abbreviations Used CI ¼ confidence interval CT ¼ computed tomography GI ¼ gastrointestinal GU ¼ genitourinary MRI ¼ magnetic resonance imaging PCN ¼ percutaneous nephrostomy SD ¼ standard deviation UTI ¼ urinary tract infection UU ¼ ureteral unit
This article has been cited by: 1. Ming-Hong Kao, Chung-Cheng Wang. 2014. The efficacy and safety of ureteral dilation and long-term type ureteral stent for patients with ureteral obstruction. Urological Science . [CrossRef]
