- Top of page
- PATIENTS AND METHODS
- CONFLICT OF INTEREST
Upper tract urothelial carcinoma (UC) is a relatively uncommon tumour, comprising only 5% of all urothelial tract tumours and 10% of renal tumours . Until recently the standard treatment for these tumours was open radical nephroureterectomy (NU), with 10–30% of patients developing metastases and an overall disease-free survival of 45–90% at 5 years [2–4].
With the emergence of improved endoscopic technology, it became feasible for upper tract UC to be diagnosed and managed ureteroscopically or percutaneously. The endoscopic approach was initially limited to patients with solitary kidneys, bilateral disease, or renal insufficiency. However, the successes of these early endeavours have encouraged practitioners to use this approach in patients with unilateral disease and normal contralateral kidney function. While the morbidity of endoscopic procedures is lower than for radical treatment, many still have reservations about using endoscopy as a primary treatment for upper tract UC. One difficulty with this treatment is that it involves intensive surveillance, requiring ureteroscopy every 3–6 months, as the sensitivity of alternative means of surveillance, cytology and imaging studies, is much lower [5,6]. Furthermore, staging with ureteroscopic biopsy is difficult and often uncertain [5,6]. Another potential disadvantage is the theoretical risk of causing metastases via pyelovenous or pyelolymphatic backflow or nephrostomy-tract seeding, although none of these consequences has yet been supported by published reports [6,7].
In the present study, we report on patients treated for upper tract UC from 1990 to 2005 at the University of Texas Southwestern department of urology, comparing the outcomes of those treated with NU immediately after diagnosis to those treated initially by conservative means.
PATIENTS AND METHODS
- Top of page
- PATIENTS AND METHODS
- CONFLICT OF INTEREST
After obtaining approval by the local Institutional Review Board, we retrospectively reviewed consecutive patients treated for upper tract UC from August 1990 to August 2005; the study included 108 patients and 120 renal collecting systems. For statistical analysis, patients were divided into two groups according to their management strategy. The immediate NU group comprised patients who had NU with no endoscopic biopsy and those who had NU within 3 months of the initial endoscopic biopsy. The conservative group comprised patients managed endoscopically and those treated with NU ≥3 months after the initial endoscopic biopsy (termed ‘delayed NU’). The surgical technique for NU included open NU with excision of a bladder cuff, or open distal ureterectomy (31 patients), laparoscopic NU (14), or laparoscopic NU with open excision of the bladder cuff (37). The selection of management strategy and surgical technique was at the discretion of the individual surgeon and patient, but was driven by the presence of bilateral disease, solitary kidney, and patient comorbidities. Conservative approaches were more prevalent in patients with low-grade disease, whereas early radical procedures prevailed in patients with high-grade disease. Thus, each management strategy was analysed according to the initial grade, which was determined by ureteroscopy when possible, although this remained indeterminate in four patients. If initial biopsies were indeterminate subsequent biopsies/washings were used.
All patients were followed with serial radiographic imaging, cytology and cystoscopy. Those patients managed conservatively also had surveillance ureteroscopy at 3–6-month intervals. During endoscopy, tumours were either biopsied with a wire basket or forceps, ablated with a holmium laser, or both.
The association between categorical data was tested using the chi-square test or Kruskal–Wallis test. Differences in continuous variables across nonparametric variables were tested using the Mann–Whitney U-test. The disease-specific survival (DSS) was compared between the groups using Kaplan-Meier survival curves and log-rank analysis. Statistical significance for all tests was set at P < 0.05 and all reported P values were two-sided.
- Top of page
- PATIENTS AND METHODS
- CONFLICT OF INTEREST
In all, 108 patients and 120 renal collecting systems were treated for upper tract UC, including 75 (69%) men and 33 (31%) women. The most common indications for upper tract evaluation were haematuria (54.9%), surveillance in patients with bladder cancer (22.1%) and flank pain (8%). In seven of 108 (6.5%) patients the upper tract lesion was discovered incidentally. Immediate NU was used in 79 renal units (65.8%), while conservative therapy was used in 41 (34.2%). Within the conservative group, delayed radical surgery was used in 11 renal units, with a median (range) time to NU of 22 (4–64) months, while 30 renal units were managed only endoscopically.
Overall, 76 patients (70.4%) were alive at a median (range) follow-up of 45.8 (0.5–129.0) months, while 16 (14.8%) died from unrelated causes, and 16 (14.8%) died from disease. The median follow-up for the renal units of patients alive at the last follow-up was 46 and 33 months for the immediate NU and conservative groups, respectively. The mean (sd) 5-year overall survival (OS) in the immediate NU and conservative groups was 72.1 (6.0)% and 61.7 (10.0)% (P = 0.356), and the 5-year DSS was 83.0 (5.1)% and 81.6 (8.5)% (P = 0.980).
There were 48 low-grade tumours, determined by initial ureteroscopic biopsy or subsequent biopsies, of which 27 (56%) were managed conservatively and 21 (44%) by immediate NU (Table 1). The median age at first surgery for patients managed conservatively or by immediate NU were 68 (33–89) and 65 (51–88) years, respectively. In patients with low-grade disease, there were 27 right-sided and 21 left-sided tumours. There were 14 tumours with concurrent contralateral tumours, seven solitary kidneys, and 27 with a normal contralateral kidney. Two of 17 patients with low-grade disease confirmed on initial pathology in the conservative treatment group had tumours that were nodular in appearance, while three of 15 low-grade tumours in the immediate NU group had tumours with a nodular appearance (P = 0.106).
Table 1. A summary of treatment regimens in low- and high-grade tumours
|Group||Immediate NU||Conservative treatment||P*|
| No. of patients||21||27|| |
| Median (IQR) age, years||65.0 (19)||68.0 (17)||0.852|
| Bilateral disease|| 4||10||0.214|
| Solitary kidney|| 1|| 6||0.118|
| Nodular appearance|| 3|| 2||0.102|
| Median (IQR) follow-up, months||37 (27.5)||29 (52.5)||0.564|
| 5-year OS, % (sd)||66.4 (12.6)||75.4 (12.8)||0.281|
| 5-year DFS, % (sd)||87.4 (8.4)||86.2 (9.1)||0.909|
| No. of patients||56||12|| |
| Median (IQR) age, years||65 (15)||71.5 (12)||0.203|
| Bilateral disease|| 6|| 3||0.190|
| Solitary kidney|| 3|| 7||<0.001|
| Median (IQR) follow-up, months||48.5 (44.5)||53.5 (38.8)||0.802|
| 5-year OS, % (sd)||71.5 (16.6)||45.0 (16.6)||0.077|
| 5-year DSS, % (sd)||75.0 (8.1)||68.6 (18.6)||0.528|
The median follow-up in patients alive in both groups was not significantly different (P = 0.564; Table 1). The OS between conservative treatment and immediate NU were similar for low-grade disease at 5 years, with a mean (sd) 75.4 (12.8)% and 66.4 (12.6)% (P = 0.281; Fig. 1a). The DSS at 5 years in patients with low-grade disease was equally good for conservative treatment and immediate NU, at 86.2 (9.1)% vs 87.4 (8.4)% (P = 0.909; Fig. 1b).
Figure 1. Comparison of the OS (a,c) and DSS (b,d) in patients with low-grade (a,b) or high-grade (c,d) UC undergoing immediate NU vs conservative initial management.
Download figure to PowerPoint
In the conservative group, seven patients required delayed NU due to stage or grade progression. The mean (sd) recurrence was 1.26 (2.01) per renal unit. The grade progressed in four renal units and stage progressed in two, including one with progression of both. In all, 94 endoscopic procedures were performed during the follow-up.
Bladder cancer was diagnosed in 28 renal units with low-grade upper tract UC. There were 15 diagnoses of bladder cancer before diagnosis of upper tract UC, of which 11 were in the conservative and four in the immediate NU group. Three bladder cancers were diagnosed at the time of the upper tract UC diagnosis (one in the conservative and two in the immediate NU group), and 10 bladder cancers were diagnosed after treatment for upper tract UC (four in the conservative and six in the immediate NU group).
There were 68 high-grade tumours, of which 12 (18%) patients had conservative management and 56 (82%) had immediate NU (Table 1). Thirty of these tumours were right-sided and 38 were left-sided. The median ages for the conservative and immediate NU groups were 71.5 (47–81) and 65.0 (35–86) years, respectively (P = 0.203). Five patients treated conservatively and 13 treated with immediate NU had carcinoma in situ. Among those treated conservatively, seven of 12 had a solitary kidney and three had bilateral disease, while three of 56 treated with immediate NU had solitary kidneys and six had bilateral disease. Five patients treated conservatively had other prohibitive factors, including metastatic RCC (one), lung cancer (one), severe heart disease and chronic obstructive pulmonary disease (two), and a final patient was stable after a cardiac stent and chose NU at 4 months.
The median follow-up was not significantly different between the groups (P = 0.820, Table 1, Fig. 1c). The OS at 5 years for the conservative and immediate NU groups were 45.0 (16.6)% and 71.5 (16.6)% (P = 0.077) and the respective DSS were 68.6 (18.6)% and 75.0 (8.1)% (P = 0.528; Fig. 1d).
In the conservative group, four patients required delayed NU; 37 endoscopic procedures were performed. The mean recurrence rate was 1.40 (2.80) per patient (Table 2). There were 14 recurrences in four patients and progression of stage in three. Due to their comorbidities, two patients did not have repeat procedures, thus their true endoscopic recurrence is unknown.
Table 2. A summary of the treatment regimen for the conservative group
|No. of patients||27||12|| |
|Endoscopic procedures||94||37|| |
|Mean (sd)/patient|| 3.48 (3.02)|| 3.25 (4.73)||0.854|
|Recurrence, n (%) patients||13 (48)|| 4||0.725|
|Mean (sd)/patient|| 1.26 (2.01)|| 1.40 (2.80)||0.866|
| Mean (sd) time, months||19.4 (17.4)||26.0 (32.1)||0.713|
| Increased stage|| 2*|| 3||0.091|
| Increased grade|| 4*|| 0|| |
Bladder cancer was diagnosed in 38 patients with high-grade upper tract UC. There were 24 diagnoses of bladder cancer before diagnosis of upper tract UC, of which six were in the conservative group and 18 in the immediate NU group. Six patients were diagnosed with bladder cancer at the time of upper tract UC diagnosis (all in the immediate NU group), and eight bladder cancers were diagnosed after the diagnosis of upper tract UC (two in the conservative and six in the immediate NU group).
The final pathology at the time of NU or ureterectomy for the immediate and delayed NU groups is shown in Table 3. There were more patients with stage ≥2 disease in the delayed NU group than in the immediate NU group (four of 11 vs 23%). Statistical significance was not definitive, as there were too few patients having delayed NU.
Table 3. Pathological comparison of local stage at NU between renal units with early and delayed NU
|Variable||Immediate NU||Delayed NU||P|
|No. of patients||79|| 11|| |
| Not available|| 1|| 1|| |
| Benign|| 2|| 0|| |
| Ta||28|| 5|| |
| CIS only|| 2|| 0|| |
| T1||27|| 1|| |
| ≥T2||18|| 4||0.830*|
| Not available|| 3|| 1|| |
| Low||19|| 2|| |
| High||57|| 8||0.787†|
There was disease recurrence in 35 (44%) of 79 renal units treated with immediate NU, vs eight of 11 treated with delayed NU (P = 0.075). The location of recurrent disease was the bladder, local or distant in 18, five and 12 renal units in the early NU vs three, two and three in the delayed NU group, respectively (P > 0.05 for all types of recurrence). In six of the seven local recurrences, recurrence was after a distal ureterectomy. Metastasis was found in five patients treated with early NU vs one treated with delayed NU.
- Top of page
- PATIENTS AND METHODS
- CONFLICT OF INTEREST
The advances in endoscopic surgery have provided an alternative to NU for treating upper tract UC. In the present analysis we examined the difference in outcomes among patients treated conservatively (initially endoscopic with or without delayed NU) compared with those undergoing early NU. There were no differences in the OS or DSS among the groups in those with low-grade disease. Although not statistically significant, patients with high-grade disease tended to have a better OS with immediate NU than those managed conservatively, but there was no trend for DSS. The 82% 5-year DSS of the conservative group is comparable with that reported in other series [8–11]. In a large retrospective review of 61 patients, Deligne et al. reported a DSS of 84%, with an OS of 77%. In the series by Elliott et al. of 44 patients undergoing endoscopic resection of upper tract UC, the 5-year DSS was 73%. Chen and Bagley  reported a survival rate of 100% at a mean of 3 years of follow-up in a prospective series of 23 patients with a normal contralateral kidney. However, 96% of the patients in that series had low-grade disease, which emphasizes the influence of grading on patient outcome.
Although previous reviews of endoscopically treated upper tract UC have raised the concern that this approach might affect pathological stage at the time of NU in those who require NU, this was not supported by our study. Progression in stage from serial ureteroscopic biopsy to NU or repeat ureteroscopy in this series was low, occurring in 12% of renal units managed endoscopically. Progression in grade was similarly low (12%). This low rate of progression was also noted by Keeley et al. in 92 patients undergoing endoscopic treatment (38 patients) vs immediate NU, in which no instances of progression were reported. In a series by Martinez-Pineiro et al. two of 28 patients had disease progression. Elliott et al. reported six patients with disease progression of 44 managed endoscopically, of whom five had muscle-invasive bladder cancer and were unable to undergo cystectomy at the time of the diagnosis of upper tract UC.
The pathological stage distribution at the time of delayed NU was similar to that in patients undergoing immediate NU. However, the proportion of patients with high-grade disease at the time of delayed NU was higher than in those who had immediate NU. This was probably because patients with initial low-grade disease were more likely to be managed conservatively until they progressed in grade or stage. Other series reported little effect of endoscopic diagnosis or management on the final pathological outcome [7,14]. In a series of 121 patients having NU, Boorjian et al. reported comparable pathological distributions between patients not having previous ureteroscopy, those having only one diagnostic biopsy, and those with ureteroscopic treatment before eventual NU. They also found a similar overall disease-free survival rate in patients undergoing immediate or delayed NU. Another study by Hendin et al. compared 48 patients undergoing NU without previous endoscopy with 48 undergoing previous biopsy, and found no significant difference in the rate of recurrence, metastases, or OS over 5 years (87% and 76%, respectively). Further research on patients undergoing delayed NU is required to determine its affect on both pathological and overall outcome.
One of the criticisms of endoscopic treatment of upper tract UC is the potential for a higher rate of subsequent bladder cancer. More aggressive surgical approaches are reportedly associated with a lower overall bladder recurrence rate, of 23–30%, in a review of open series [2,3]. In addition to a longer period of exposure of the bladder to an upper tract UC, the bladder could be seeded during the ureteroscopic removal of the upper tract tumour. However, in a series by Keeley et al., a third of patients developed de novo tumours, which is similar to most series in which patients had immediate NU. In the present series, despite the high rate of bladder cancers before the diagnosis of upper tract UC, which was significantly higher in the conservative group (47% vs 27%), the rate of de novo bladder cancer appearing after diagnosis was low in each group (13% in the conservative vs 19% in the immediate NU group). Thus, the risk of developing subsequent bladder cancer after managing upper tract UC is unlikely to be related to the type of approach.
There are several limitations of the present analysis. One important limitation is the reliance on pathological grading and staging by ureteroscopy in those managed conservatively, to document progression. Endoscopic staging can be unreliable. However, a correlation between grade and stage on ureteroscopic biopsy was reported [5,6]. For instance, Keeley et al. found only 10% of low- or moderate-grade tumours on biopsy with invasive disease on NU, vs 11 of 12 with high-grade tumours. Relative to other means of diagnosis or surveillance, Chen et al. reported an overall sensitivity and specificity of 93.4% and 65.2%, respectively, for ureteroscopic biopsies or cytology. The sensitivity of urine analysis, bladder washings and retrograde pyelography were much lower (37.5%, 50% and 71.7%, respectively) . Thus, ureteroscopy is the most accurate means available for following these patients.
Another limitation of this analysis is the presence of concomitant bladder cancer in the groups. Overall, there was a high frequency of concomitant bladder cancer, in nearly two-thirds of the patients, making an evaluation of the source of disease recurrence difficult. Moreover, the presence of bladder cancer can affect the calculated survival outcomes. This dilemma is shared by others evaluating upper tract UC, where the rate of concomitant bladder cancer is 16–84%[1,8,10,12]. In addition, the presence of bladder cancer before the diagnosis of upper tract UC, even if treated, affects the recurrence of UC in the bladder and upper tract . In a series of 44 patients, 53% had bladder cancer before the diagnosis of upper tract UC . During the follow-up, 45% of patients with a history of bladder cancer developed recurrent bladder cancer, while only 33% of those with no such history developed de novo bladder cancers . In the present series, only two of 16 patients, who died from UC, had muscle-invasive bladder TCC or had a cystectomy for bladder cancer. Thus, although recurrence is affected by concomitant bladder cancer, the impact of invasive bladder cancer on OS in this group is low.
Finally, conclusions drawn from this analysis are limited by its retrospective design and relatively few patients. Comparison of more patients who have delayed NU after a period of surveillance will be an important follow-up study. Ideally, survival could be compared between delayed and immediate NU groups from the time of NU. This analysis is not a comparison of efficacy of the two types of treatment (conservative vs immediate NU); there is a selection bias and tumours that are larger and bulkier were more likely to be selected for radical NU from the start. This study shows that conservative management for low-grade tumours had a comparable outcome to initial radical treatment. Also, conservative management followed by radical NU when necessary for tumour stage/grade progression did not appear to compromise the outcome.
In conclusion, in this series we compared the outcomes between patients undergoing immediate NU and those treated conservatively, showing a similar DSS in those with low-grade disease. In all, 31 renal units of a possible 41 were preserved, suggesting that endoscopic management of UC can be used to avoid surgery in patients who might be at high risk of surgical morbidity. However, those with high-grade disease tend to do poorly and should be treated by aggressive means whenever possible.