Ureteroscopic and extirpative treatment of upper urinary tract urothelial carcinoma: a 15-year comprehensive review of 160 consecutive patients

Authors


Michael Grasso, 100 East 77th Street, Suite 4 East Building, New York, NY 10075, USA. e-mail: Mgrasso3@earthlink.net

Abstract

Study Type – Therapy (case series)

Level of Evidence 4

What's known on the subject? and What does the study add?

Upper urinary tract urothelial carcinomas (UTUC) have historically been treated with radical, extirpative surgery, primarily nephroureterectomy with bladder-cuff excision. In general, there has been growing interest in renal preservation, as evidenced by the broadening application of nephron-sparing surgery for renal parenchymal tumours. Beyond imperative reasons such as tumour in a solitary kidney, bilateral disease, or comorbidities preventing radical surgery, there is a growing role for endoscopic management of upper tract tumours. The aim has been to obtain similar oncological results to those of extirpative surgery, while preserving long-term renal function. Properly selecting patients for these therapies, designing specific treatments based on a complex presentation, and general information with regard to outcomes and risks for patient counselling have been based historically on results from relatively small series without long-term follow-up.

This study reflects all patients with UTUC treated by a single tertiary referral surgeon, accrued prospectively over 15 years using the same surgical techniques and treatment algorithms throughout the entire study period, with 10-year survival data. The consecutively accrued nature and size of the study groups, uniformity in treatments, statistical review and long-term follow-up provide baseline oncological data that could help frame future study.

OBJECTIVE

  • • To present long-term oncological outcomes of all patients treated surgically for upper urinary tract urothelial carcinoma (UTUC) over a 15-year period.

PATIENTS AND METHODS

  • • All patients (N= 160) treated from January 1996 to August 2011 were prospectively studied and placed into three distinct groups after initial diagnostic ureteroscopy (URS): Group 1: low grade lesions treated with URS (n= 66); Group 2: high grade lesions palliatively treated with URS (n= 16); and Group 3: extirpative surgery (nephroureterectomy [NU]; n= 80).
  • • Statistical analysis was performed using Kaplan–Meier methodology to calculate overall (OS), cancer-specific (CSS) and metastasis-free survival (MFS).

RESULTS

  • • The median patient age at presentation was 73 years, and the mean (range) follow-up time was 38.2 (1–185) months. At initial diagnostic URS, 71 (44.4%) patients presented with high grade and 89 (55.6%) patients presented with low grade disease.
  • • The 2-, 5- and 10-year CSS rates were 98, 87 and 81% for patients with low grade disease, and 97, 87 and 78% for patients treated with URS (Group 1), not significantly different from those patients with low grade disease treated with NU (Group 3), (P= 0.54).
  • • Of the patients treated with URS for low grade disease, 10 (15.2%) progressed to high grade disease at a mean time of 38.5 months.
  • • Patients with high grade disease treated with NU had a 2-, 5-, and 10-year CSS of 70, 53 and 38%, with a MFS of 55, 45 and 35%.
  • • Median survival of patients with high grade disease treated with palliative URS was 29.2 months with a 2-year OS of 54%.
  • • On multivariate analysis only high grade lesion on initial presentation was found to be a significant factor (P < 0.001; hazard ratio = 7.27).

CONCLUSIONS

  • • Grade is the most significant predictor of OS and CSS in those with UTUC, regardless of treatment method.
  • • Ureteroscopic and extirpative therapy are acceptable options for those with low grade disease showing excellent long-term CSS.
  • • Extirpative therapy was found to result in relatively poor long-term CSS in patients with high grade disease, underscoring the need for adjuvant or neoadjuvant therapies.
Abbreviations
UTUC

upper urinary tract urothelial carcinoma

URS

ureteroscopy

NU

nephroureterectomy

CSS

cancer-specific survival

MFS

metastasis-free survival

OS

overall survival

CCI

Charleston comorbitiy index

HR

hazard ratio.

INTRODUCTION

Upper urinary tract urothelial carcinomas (UTUCs) have historically been treated with radical, extirpative surgery, primarily nephroureterectomy (NU) with bladder-cuff excision. This is particularly true for large, bulky lesions and those associated with upper urinary tract obstruction. There has been growing interest in renal preservation, as evidenced by the broadening application of nephron-sparing surgery for renal parenchymal tumours. Similarly to the endoscopic treatment of lower urinary tract urothelial malignancies, techniques in renal preservation with regard to UTUC have been based on ureteroscopic therapies.

Beyond imperative reasons, such as tumour in a solitary kidney, bilateral disease, or comorbidities preventing radical surgery, there is a growing role for the endoscopic management of upper tract tumours. The aim has been to obtain similar oncological results to those obtained by extirpative surgery, while preserving long-term renal function. Properly selecting patients for these therapies, designing specific treatments based on a complex presentation, and general information with regard to outcomes and risks for patient counselling have been based historically on results from relatively small series without long-term follow-up. The aim of the present study was to design and use aggressive endoscopic treatments for this disease, stratifying patients based on characteristics defined at initial diagnostic ureteroscopy (URS), and then maintaining meticulous long-term follow-up. The study began in 1996 and early results were presented in 1999 and 2005 [1,2]. In the present study we report on our 15-year cumulative experience.

PATIENTS AND METHODS

All patients referred for management of UTUC, treated from January 1996 to August 2011 by one surgeon, were included; there were no exclusions. Specifically, a prospective surgical log was kept with detailed information with regard to presenting characteristics and the specific surgical intervention used. Concurrently, outpatient follow-up data was retrospectively accrued and digested periodically. These data have been published previously [1,2]. Comorbidity was determined by using the age-adjusted Charleston comorbitiy index (CCI). The Social Security Death Index was also used. The collected information was compiled into a comprehensive institutional review board-approved database.

PATIENT POPULATION AND STUDY DESIGN

A total of 160 patients were studied. Demographic characteristics are shown in Table 1. Patients were stratified into one of three groups based on a preliminary diagnostic evaluation performed on presentation. This diagnostic evaluation included CT or MRI of the abdomen and pelvis, additional imaging as part of a metastatic evaluation, urine cytology, and diagnostic URS.

Table 1.  Patient demographics and findings at initial diagnosis
 URS groupsExtirpative surgery groupAll groups
Group 1Group 2Group 3
No. of patients661480 160
Median age (range)73 (45–93)71.5 (48–89)72.5 (46–90) 73 (45–93)
Sex, n (%)     
 Male35 (53)1150 (62.5) 96 (60)
 Female31 (47)330 (37.5) 64 (40)
 Mean age-adjusted CCI (range)5.5 (0–13)7.1 (2–11)5.1 (0–13) 5.4 (0–13)
 Initial/last follow-up mean creatinine concentration, mg/dL1.5/1.72.0/2.51.3/1.6 1.4/1.7
  Solitary kidney, n (%)17 (25.8)62 (2.5) 25 (15.6)
 Bilateral disease, n (%)9 (13.6)54 (5) 18 (11.3)
  Multifocality, n (%)21 (31.%)1123 (28.8) 55 (34.4)
Urothelial cancer, n (%)    
 Preoperative history of bladder cancer29 (43.9)1021 (26.3) 60 (37.5)
  High grade9 (13.6)712 (15) 28 (17.5)
  Low grade23 (34.8)411 (13.8) 38 (23.8)
 Concomitant bladder cancer17 (25.8)520 (25) 42 (26.3)
  History of cystectomy1 (1.5)41 (1.3) 6 (3.8)
 Preoperative history of upper tract cancer19 (28.8)76 (7.5) 32 (20)
  High grade6 (9.1)30 (0) 9 (5.6)
  Low grade13 (19.7)46 (7.5) 23 (14.4)
Initial grade at diagnosis, n (%)    
 High grade0 (0)1457 (71.3) 71 (44.4)
 Low grade66 (100)023 (28.8) 89 (55.6)
Initial upper tract tumour, n (%)    
 Large: >3 cm18 (27.3)962 (77.5) 89 (55.6)
 Medium: 1–3 cm32 (48.5)214 (17.5) 48 (30)
 Small: <1 cm16 (24.2)34 (5) 23 (14.4)
 Mean size, cm2.33.073.5 3.0
Mean (range) follow-up, months 51.5 (3–166.4)25.1 (6.5–47.8)30.4 (1–185.3) 38.2 (1–185.3)

Group 1: URS therapy group. The inclusion criteria for the URS group included a ureteroscopic biopsy confirming a low grade neoplasm, a negative upper urinary tract cytology for high grade cells, tumour size amenable to endoscopic resection, acceptable residual renal function, and the patient agreeing to a lifelong strict surveillance protocol including interval URS. Patients with a low grade biopsy and a positive cytology for high grade cells were not offered this treatment. Patients with a non-diagnostic biopsy and negative cytology underwent repeat URS for further tissue sampling.

Flexible ureteroscopic mapping was performed in all patients, with representative biopsies obtained for histopathology, and barbotage specimens for cytopathology. Ureteroscopic tumour therapy was performed with either electrocautery or laser energy, removing all visible tumour as described previously [3,4]. Staged URS was performed for large lesions. Concurrent bladder tumours were treated with standard therapies. Ureteroscopic surveillance was performed frequently at the outset (i.e. every 3 months after initial tumour resection), less frequently with negative evaluations, but with a minimum of an annual upper urinary tract endoscopy and semi-annual cystoscopy. Metastatic evaluations were performed annually. Adjuvant topical upper urinary tract therapy, with either mitomycin C or α-interferon, was used selectively in those with large-volume or multifocal lesions.

Patients in the URS therapy group who progressed in grade were offered entry into one of the other two groups, but for statistical evaluation were maintained in Group 1. When a ureteroscopic biopsy or cytology converted to high grade, these patients were directed to NU. If comorbidities prohibited extirpative surgery, patients were treated with palliative URS.

Group 2: Palliative URS group. Patients included in the palliative URS group had high grade UTUCs, and were either not candidates for NU owing to comorbidities, or refused extirpative surgery. The majority of these patients were obstructed at presententation and had significant gross haematuria.

All patients in this group would require chronic haemodialysis if NU was performed and so the treatment goal was to maintain renal function without haematuria. As with Group 1, URS was based on complete tumour resection, in a staged fashion if necessary, with subsequent surveillance. Patients were counselled from the outset that metastatic disease was probable.

Group 3: Extirpative therapy (open and laparoscopic NU) group. Patients who underwent extirpative resection of a UTUC as primary treatment were evaluated and followed. The majority of patients underwent laparoscopic NU via a retroperitoneal approach, with transurethral incision of the ureteric orifice and then open bladder-cuff excision through a Gibson incision. Open surgical intervention was reserved for those patients who had previously undergone open surgical intervention or had obvious high-stage disease. Distal ureterectomy was infrequently used (n= 2). Lymphadenectomy was routinely performed for staging. No patients received adjuvant or neoadjuvant chemotherapy as part of this protocol.

Patients were followed up with serial cystoscopic evaluations and urine cytology, with an increasing interval if the previous results were acceptable with a minimum annual evaluation. Metastatic evaluations were performed semi-annually for 2 years, and if negative then annually.

PATHOLOGY

The tumour staging system used for all patients was based on the 2002 American Joint Committee on Cancer/Union Internationale Contre le Cancer TNM system and graded according to the 2004 WHO/International Society of Urological Pathology consensus classification system.

STATISTICAL ANALYSIS

Statistical analysis was performed using Prism 5.0 and NCSS 7.1.21. Survival data were calculated using Kaplan–Meier analysis. Prognostic variables that affected overall (OS) and cancer specific survival (CSS) were established by univariate analysis using the log-rank test and further analysed by multivariate analysis using the Cox proportional hazards regression model. Variables such as age, sex, comorbidity, history of bladder cancer, tumour grade and stage, tumour size, and upper urinary tract multifocality were all analysed in the overall population and within the individual groups. All tests were two-sided and a P value <0.05 was considered to indicate statistical significance.

RESULTS

GROUP 1: URS THERAPY GROUP

Peri-operative data and outcomes from URS are shown in Table 2. Using meticulous ureteroscopic surveillance, 77% of patients were noted to have a recurrent lesion over time, the majority of which were small and easily cleared endoscopically. Mean time to recurrence was 12 months, maximum 62.3 months. No patients were lost to follow-up.

Table 2.  Perioperative data and outcomes of URS
 Group 1 (low grade)Group 2 (high grade)
No. of patients 6614
Total no. of URSs 52275
Mean no. of URSs per patient 7.95.4
Staged URS required for tumour clearance, % 49/329 (14.9)28/75 (38)
Urothelial recurrence   
 Upper tract recurrence, n5114
 Median (range) time to first recurrence, months9.9 (1.1–62.3)3 (1–8)
 New bladder cancer, n182
 Bladder recurrence, n407
Progression   
 Progression to high grade, n10NA
 Undergraded initial URS biopsy (subsequent biopsy in <6 months high grade), n3NA
 Overgraded on survelliance URS biopsy, n2NA
 Subsequent NU, n114
 Progression to renal failure, n45
Survival analysis   
 Deceased, n2012
 Median OS, months126.829.2
 Cancer-related deaths, n812
 Metastasis, n913

Progression in grade on subsequent evaluation was noted in 10 patients (15.2%), with a mean (range) time to progression of 38.5 (4.6–115) months (Fig. 1). Three patients (4.5%) had high grade biopsy or cytology within 6 months, as a result of undergrading. Excluding those who were undergraded initially, the mean (range) time to progression was 52 (20–115) months, with a mean (range) OS of 67.7 (27.3–166.4) months. Pan-urothelial disease, large tumour volume, and history of a high grade bladder carcinoma were prominent features in the group of patients that progressed (40%). Two patients had a ureteroscopic biopsy defining a subsequent high grade lesion, but on final NU specimen only a low grade process was noted, reflecting biopsy overgrading.

Figure 1.

Recurrence- and progression-free survival of patients with low grade disease treated with URS (Group 1).

Eleven patients (16.7%) in this group underwent NU either for grade progression, poor compliance, or poor renal function. The majority of patients maintained adequate renal function (94%) with long-term surveillance, but progression to end-stage renal disease was a positive predictor for subsequent progression in grade.

Long-term OS, CSS and metastasis-free survival (MFS) are shown in Fig. 2. The median OS of this group was 126.8 months. All patients in this group, who progressed in stage, represent only those who progressed in grade. In addition, all eight (12.1%) cancer-related deaths were in those patients who progressed with high grade disease.

Figure 2.

OS, CSS and MFS, stratified by treatment groups.

GROUP 2: PALLIATIVE URS GROUP

Outcomes from the palliative URS group are shown in Table 2. The majority of patients in this group initially presented in extremis from bleeding and/or renal insufficiency. Adequate local control was obtained in all patients with resolution of gross haematuria. Patients in this palliative ureteroscopic therapy group had urothelial recurrences subsequently treated. In addition, the majority (N = 15) were found to have metastatic disease on subsequent evaluations (93%). Five patients progressed to end-stage renal disease in this group, four of whom underwent NU thereafter. Survival data are shown in Fig. 2, with a median (range) OS of 29.2 (6.5–52.6) months.

GROUP 3. EXTIRPATIVE THERAPY (OPEN AND LAPAROSCOPIC NU) GROUP

The results of those treated primarily with extirpative surgery, and those treated with subsequent NU, are shown in Table 3. The majority of tumours in this group were invasive and high grade (71%). CSS rates after primary extirpative surgery are shown in Fig. 3. Ten (12.5%) had positive regional lymph nodes. As expected, a higher stage portended shorter survival, with a median CSS of 23, 5.7, 16.5 months for patients with stages T3, T4, and patients with positive lymph nodes, respectively.

Table 3.  Perioperative data and outcomes of extirpative surgery
 Group 3All Patients
No. of patients 8095
Surgical treatment, n (%)   
 Laparoscopic NU72 (90)82 (86.3)
 Open NU6 (7.5)10 (10.5)
 Segmental ureterectomy2 (2.5)3 (3.2)
Tumour stage, n (%)   
 Tis1 (1.3)5 (5.3)
 Ta28 (35)31 (32.6)
 T112 (15)13 (13.7)
 T213 (16.3)14 (14.7)
 T322 (27.5)26 (27.4)
 T44 (5)6 (6.3)
 Lymph-node-positive10 (12.5)16 (16.8)
Pathological grade, n (%)   
 High57 (71.3)69 (72.6)
 Low23 (28.8)26 (27.4)
Urothelial recurrence, n (%)   
 Upper tract contralateral recurrence4 (5)7 (7.4)
 New bladder cancer13 (16.3)20 (21.1)
 Bladder recurrence22 (27.5)35 (36.8)
Survival analysis   
 Deceased, n (%)30 (37.5)39 (41.1)
 Median OS, months109.2NA
 Cancer-related deaths, n (%)24 (30)33 (34.7)
 Metastasis, n (%)30 (37.5)41 (43.2)
  High grade metastasis, n (%)28 (93.3)39 (95.1)
  Low grade metastasis, n (%)2 (6.7)2 (4.9)
Figure 3.

CSS of patients treated with primary extirpative surgery stratified by stage and lymph node status.

Metastatic disease was found in 38% of patients at a mean time of 13.5 months. Only two patients had low grade metastasis, both with very large intrarenal tumours and tiny multifocal pulmonary nodules which, on subsequent biopsy, had identical histology. Both patients ultimately died from the disease within 2 years.

Figure 2 shows the OS, CSS and MFS rates. The median OS time for this group was 109.2 months, with a median OS and CSS time of 45.5 months and 72.8 months, respectively for those with high grade disease.

SURVIVAL DATA

Table 4 shows OS, CSS and MFS outcomes at 2-, 5- and 10-year follow-up, respectively. CSS of all patients, who initially presented with low grade disease, at 2-, 5-, and 10-year follow-up was 98, 87 and 81% (Fig. 4). URS for low grade tumours (Group 1) had a 2-, 5-, and 10-year CSS of 97, 87, and 78%, not significantly different from those patients with low grade tumours treated with NU (P= 0.54 [Fig. 5]).

Table 4.  OS, CSS and MFS outcomes
 2-year, %5-year, %10-year, %
OSCSSMFSOSCSSMFSOSCSSMFS
All groups, N= 160 828476596764405958
Initial grade at URS presentation          
 Low959895768785598181
 High64655033353302323
Ureteroscopic management          
 Group 1 (low grade): URS939797748784567875
 Group 2 (high grade): palliative URS545434000000
Extirpative surgery          
 (Group 3) (NU)777866586460365654
 Pathological stage         
  Ta-Tis-T1919486788678528074
   Low grade10010095889395739395
   High grade83887868786105851
  T2919174546765546765
  T349493222220000
  T4252502525025250
 Lymph-node-positive2424NA00NA00NA
 Pathological grade on NU         
  Low10010091798991668991
  High67705547534503835
Figure 4.

CSS and MFS of all patients, based on initial tumour grade at presentation (P < 0.001; HR = 7.14).

Figure 5.

CSS of all patients with low grade disease stratified by treatment method (P =0.54).

Patients who presented with high grade disease had a 2-, 5-, and 10-year CSS of 65, 35 and 23%, respectively (Fig. 4). High grade tumours, treated primarily with NU, had a slightly better 2-, 5-and 10-year CSS of 70, 53 and 38%, rates which reflect the exclusion of the palliative URS group.

UNIVARIATE AND MULTIVARIATE ANALYSIS

Predictive variables of CSS, using univariate and multivariate analysis, are shown in Table 5. Based on univariate analysis, age-adjusted CCI ≥6, preoperative bladder carcinoma and recurrence, high grade lesion, large tumours, and progression to end-stage renal disease were significant negative variables; however, based on multivariate analysis, only high grade lesions at presentation was a statistically significant variable (P < 0.001; hazard ratio = 7.27; CI 3.25–15.7).

Table 5.  Predictive variables of OS and CSS using univariate and multivariate analysis
Prognostic variableOSCSS
Univariate analysisMultivariate analysisUnivariate analysisMultivariate analysis
HR95% CI P HR95% CI P HR95% CI P HR95% CI P
  • *

    Significant at P < 0.05 (statistically significant values on univariate analysis were selected for multivariate analysis).

Age (≥73 years)1.480.90–2.450.126   1.190.66–2.140.571   
Sex (M/F)1.310.79–2.170.314   1.570.87–2.860.157   
Age-adjusted CCI (≥6) 2.28 1.36–3.82 0.002* 2.01 1.16–3.46 0.012* 2.03 1.11–3.71 0.021* 1.88 0.98–3.6 0.056
Preoperative bladder cancer (Yes/No) 1.87 1.09–3.22 0.627 1.390.78–2.470.27 1.87 1.00–3.51 0.034* 1.100.55–2.180.97
Concomitant bladder tumour (Yes/No)0.790.44–1.420.46   0.680.34–1.340.313   
Initial upper tract tumour grade (High/Low) 3.76 2.16–6.53 <0.0001* 3.78 2.11–6.80 <0.001* 7.27 3.86–13.71 <0.0001* 7.14 3.25–15.7 <0.0001*
Initial tumour size (>3 cm)1.560.95–2.590.082    2.29 1.27–4.13 0.008* 1.740.88–3.440.11
Bilateral disease (Yes/No)1.090.51–2.350.82   1.430.57–3.590.382   
Solitary kidney (Yes/No)0.860.44–1.680.67   0.950.43–2.090.892   
Multifocality (Yes/No)1.250.74–2.130.39    1.76 0.94–3.28 0.058 1.280.65–2.500.47
Bladder recurrence (Yes/No) 0.58 0.35–0.96 0.031* 0.650.38–1.120.120.600.33–1.090.089   
Progression to renal failure (Yes/No) 2.94 1.04–8.27 0.001* 1.440.83–3.790.14 4.04 1.25–13.06 <0.0001* 2.28 0.96–5.45 0.06

DISCUSSION

The study of UTUC is challenging because of the rare nature of the disease and the lack of treatment uniformity at different centres. Consequently, there is value in a long-term prospectively accrued database of all patients undergoing surgical treatment where goals are defined based on initial diagnostic URS. Limitations of other studies include selection bias, lack of uniformity of treatment, and the pooling of retrospective data from different sites.

The present study included all patients with UTUC treated by a single tertiary referral surgeon, accrued prospectively over 15 years, using the same surgical techniques and treatment algorithms throughout the entire study period. The placement of patients into one of three treatment groups at presentation was based on observations from other researchers that low grade lesions in the upper urinary tract, similarly to those in the bladder, may not require radical extirpative surgery. In 1945, Vest [5] and other later studies [6,7] reported early experiences with limited resection of benign-acting UTUCs. Zincke's pioneering work [8] at the Mayo clinic, 30 years ago, framed an argument against radical surgery for low grade UTUC, and led to the popularity of distal ureterectomy as treatment.

The endoscopic treatment of UTUC evolved in the 1980s as ureteroscopes improved [9,10]. Bagley's work on the study of diagnostic and therapeutic URS has fundamentally changed how we treat urothelial lesions in the upper urinary tract [9,11–13]. Advances in this technology include the actively deflectable, flexible ureteroscope allowing a complete evaluation of the intra-renal collecting system, and various laser light energies delivered through flexible quartz fibres sufficiently powerful and precise to facilitate complete tumour ablation [14]. In addition, experts in this field included Segura at the Mayo clinic [15,16] and Streem et al. [17] in Cleveland, who kept similar databases and whose findings underscored the usefulness of endoscopic therapy in selected patients with UTUC.

The potential limitations of the ureteroscopic treatment of UTUC include undergrading on initial biopsy and the inability to completely evaluate the upper urinary tract urothelium. Similarly to that which is found with cystoscopic mapping for bladder carcinoma, we observed that there was a small cohort (N = 3) in the URS therapy group who were undergraded on initial ureteroscopic biopsy (4.5%), thus emphasizing the need for serial complete endoscopic evaluations [18]. It is important to note that tissue sampling reviewed for histopathology, and barbotage specimens obtained through the endoscope and evaluated for cytopathology, are both essential in defining the tumour grade on presentation [12,19,20]. In addition, with the current small diameter flexible ureteroscopes, the ability to perform a complete upper urinary tract inspection during initial diagnosis and surveillance is a standard procedure widely accepted [21–23].

It is clear that urothelial tumours act differently depending on their grade. Not only is this a basic tenet of bladder cancer therapy, but is also well accepted for UTUC [23]. Gadzinski et al. [24] found at 5-year follow-up that CSS and MFS were more associated with grade at presentation than treatment choice. If we are able to completely clear low grade lesions from the upper urinary tract using URS, then we should ask, ‘What is the incidence of subsequent recurrence and progression?’ In the present series, similar to observations in those patients with low grade urothelial carcinoma in the bladder, three quarters of the patients had new low grade lesions treated [25]. Most were small and were easily removed endoscopically. With regard to survival, defining the rate of progression in grade is therefore of fundamental importance. In the present series, 15% of those treated with URS progressed in grade, with a mean time to progression of 52 months. Large and pan-urothelial tumour burden with development of end-stage renal disease portend progression and ultimate death.

It is important to place into context these patients treated with ureteroscopic resection of low grade urothelial lesions. It has been well documented that elderly patients, once placed on haemodialysis, have limited survival, as shown by 5-year OS rates of ≈19% and 10% for patients >65 and 75 years old, respectively [26]. Given that the rate of bilateral disease, solitary kidney, and modest renal insufficiency with an age-adjusted CCI >6 was significant in the present study, most of these patients would probably have required haemodialysis and may have had a lower OS if a NU had been performed instead of URS. By contrast, ureteroscopic treatment in the present study maintained renal function in 94% of patients, with CSS rates similar to all patients presenting with low grade disease, irrespective of treatment type.

The ‘gold standard’ treatment for UTUC is considered to be NU, so it is important to study outcomes in patients who undergo this treatment. Berger et al. [27] from the Cleveland Clinic presented a series of 100 NUs, with an 8-year follow-up. Their initial CFS rates were marginally better than those in the present series, while long-term OS rates were identical. These early differences reflect a greater proportion of patients with high grade (71 vs 58%), high stage UTUC, and significantly more lymph-node-positive disease (12.5 vs 4%) in the present series. Nevertheless, the MFS rates were almost identical in the two studies: 66, 50, and 36% at 2, 5 and 7 years for Gill vs 66, 60 and 54% at 2, 5, and 10 years in the present study. With a low overall MFS at 5 and 10 years, it is clear that either neoadjuvant or adjuvant therapy is needed in patients with high grade disease treated with NU [28,29].

The present data help clarify fundamental issues for those patients presenting with UTUC. High grade disease on initial URS is by far the most important prognostic indicator, with a hazard ratio (HR) of 7.12 on multivariate analysis. At 10 years the CSS of those with high grade disease treated with radical extirpative surgery was 38% compared with 89% for patients with low grade lesions (P < 0.001). The CSS rates for low grade disease at 5 years treated with URS and NU are similar (87 vs 89%), with an increasing but not significant difference in CSS at 10 years which reflects those patients who progressed in grade. The risk of progression in grade should therefore be the central component of patient counselling before embarking on URS as treatment.

Based on the present results, patients should be counselled that there is a 15% chance of progression in grade during surveillance, and thus serial URS should be an integral part of follow-up. Patients must also be aware that progression in grade commonly reflects progression in stage, with overall poor outcomes in that subset of patients. As highlighted by Keeley et al. [12,30], large pan-urothelial lesions, even though low grade on biopsy, have a higher rate of progression. That said, and similar to bladder cancer, in those patients with large low grade lesions and multiple comorbidities who would not tolerate lifelong haemodialysis, staged ureteroscopic resection can be performed successfully for local control.

A limitation of the present study is the non-randomized nature of its design in selecting treatment methods; however, randomization is difficult because of the relatively rare nature of UTUC, In addition, a single surgeon experience obviously has limitations with regard to reproducibility. Nevertheless, the consecutively accrued nature and size of the study groups, uniformity in treatments, and long-term follow-up provide baseline oncological data that could help frame future study.

In conclusion, grade is the most significant predictor of OS and CSS in those with UTUC, regardless of treatment method. Ureteroscopic and extirpative therapy are acceptable options for those with low grade disease. Extirpative therapy has relatively poor long-term CSS in patients with high grade disease, underscoring the need for possible adjuvant or neoadjuvant therapies.

ACKNOWLEDGEMENTS

We would like to thank Dr Carl V. Hamby PhD, Associate Professor of Microbiology and Immunology at New York Medical College, Valhalla, NY, USA for his assistance with the statistical analysis and survival data.

CONFLICT OF INTEREST

None declared.

Ancillary