Comparison of oncological outcomes after segmental ureterectomy or radical nephroureterectomy in urothelial carcinomas of the upper urinary tract: results from a large French multicentre study

Authors


Pierre Colin, CHRU Lille, Univ Lille Nord de France – Academic Department of Urology, LILLE, France. e-mail: pierre_colin@msn.com

Abstract

Study Type – Therapy (multi-centre retrospective cohort)

Level of Evidence 2b

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

Upper urinary tract urothelial carcinomas (UUT-UCs) are rare tumours. Because of the aggressive pattern of UC, radical nephroureterectomy (RNU) with bladder cuff removal remains the ‘gold-standard’ treatment. However, conservative strategies, such as segmental ureterectomy (SU) or endourological management, have also been developed in patients with imperative indications. Some teams are now advocating the use of conservative management more commonly in cases of elective indications of UUT-UCs. Due to the paucity of cases of UUT-UC, only limited data are available on the oncological outcomes afforded by conservative management.

We retrospectively investigated the oncological outcomes after SU and RNU in a large multi-institutional database. Overall, 52 patients were treated with SU and 416 with RNU. There was no statistical difference between the RNU and SU groups for the 5-year probability of cancer-specific survival, recurrence-free survival and metastasis-free survival. The type of surgery was not a significant prognostic factor in univariate analysis. The results were the same in a subgroup analysis of only unifocal tumours of the distal ureter with a diameter of <2 cm and of low stage (≤T2). Our results suggest that oncological outcomes after conservative treatment with SU are comparable to RNU for the management of UUT-UC in select cases.

OBJECTIVE

  • • To compare recurrence-free survival (RFS), metastasis-free survival (MFS) and cancer-specific survival (CSS) after segmental ureterectomy (SU) vs radical nephroureterectomy (RNU) for urothelial carcinoma (UC) of the upper urinary tract (UUT-UC) located in the ureter.

PATIENTS AND METHODS

  • • We performed a multi-institutional retrospective review of patients with UUT-UC who had undergone RNU or SU between 1995 and 2010.
  • • Type of surgery, Tumour-Node-Metastasis status, tumour grade, lymphovascular invasion and positive surgical margin were tested as prognostic factors for survival.

RESULTS

  • • In all, 52 patients were treated with SU and 416 with RNU. The median (range) follow-up was 26 (10–48) months.
  • • The 5-year probability of CSS, RFS and MFS for SU and RNU were 87.9% and 86.3%, respectively (P= 0.99); 37% and 47.9%, respectively (P= 0.48); 81.9% and 85.4%, respectively (P= 0.51).
  • • In univariable analysis, type of surgery (SU vs RNU) failed to affect CSS, RFS and MFS (P= 0.94, 0.42 and 0.53, respectively).
  • • In multivariable analyses, pT stage and pN stage achieved independent predictor status for CSS (P= 0.005 and 0.007, respectively); the positive surgical margin and pT stage were independent prognostic factors of RFS and MFS (P= 0.001, 0.04, 0.009 and 0.001, respectively).
  • • The main limitation of the study is its retrospective design, which is due to the rarity of the disease.

CONCLUSIONS

  • • Short-term oncological outcomes after conservative treatment with SU are comparable to RNU for the management of UUT-UC in select cases and should be considered an option.
  • • In every other case, RNU still represents the ‘gold standard’ for the treatment of UUT-UC.
Abbreviations
RFS

recurrence-free survival

MFS

metastasis-free survival

CSS

cancer-specific survival

SU

segmental ureterectomy

RNU

radical nephroureterectomy

(UUT)-UC

(upper urinary tract) urothelial carcinoma

LVI

lymphovascular invasion

INTRODUCTION

Upper urinary tract urothelial carcinomas (UUT-UCs) are rare tumours with an estimated incidence of 1–4 cases per 100 000 individuals per year. They represent only 5–6% of all UCs. Unlike bladder tumours that can be managed with endoscopic resection, sparing the organ in non–muscle-invasive cases, the ‘gold standard’ for UUT-UC treatment remains radical nephroureterectomy (RNU) [1]. The superiority of RNU is related to the presence of more locally advanced stages at presentation that are difficult to stage based on preoperative evaluation, are more complicated to reach with topical therapies and are more difficult to monitor in cases of conservative treatment [2]. However, conservative strategies, such as segmental ureterectomy (SU) or endourological management, have also been developed in patients with imperative indications (e.g. solitary kidney, synchronous bilateral disease, compromised renal function, or comorbidity precluding a radical surgical procedure) [3]. Some teams are now advocating the use of conservative management more commonly in cases of elective indications of UUT-UC [1]. Due to the paucity of cases of UUT-UC, only limited data are available on the oncological outcomes afforded by conservative management. Thus, we retrospectively investigated the oncological outcomes obtained after SU and RNU in a large multi-institutional database.

PATIENTS AND METHODS

We identified all patients who had undergone surgery (RNU or SU) for UUT-UC between January 1995 and December 2009 at 18 French institutions. Sample collection was approved by the Ethics Committee of the Assistance Publique-Hôpitaux de Paris, i.e., Institutional Review Board approval. Patients with evidence of metastatic disease at the time of diagnosis were excluded. The following data were reviewed: age at diagnosis, gender, type of surgery, tumour characteristics (TNM stage 2002, World Health Organization grade 1973), lymphovascular invasion (LVI), surgical margin status, tumour site, disease recurrence, and metastatic progression. Disease recurrence was defined as any documented failure in the bladder or in the contralateral UUT. Metastatic progression was defined as any recurrence in regional lymph nodes, including the tumour bed or distant metastasis.

For all patients, preoperative pT stratification was established by CT and/or MRI. Ureteroscopy with biopsy was not systematically practiced between 1995 and 2009, in accordance with our national guidelines. SU was indicated in imperative cases, such as patients with a solitary kidney, chronic renal insufficiency, bilateral tumour, or prohibitive operative risk (American Society of Anesthesiologists score ≥3). At the physician's discretion, SU was also offered to patients with a distal, unifocal tumour with a diameter of <2 cm, with no evidence of invasion in the preoperative evaluation.

RNU was performed by a standard double-access procedure: a loin incision followed by an iliac incision. The kidney, ureter and a bladder cuff were excised en bloc, as previously described [4]. For SU, an iliac access was achieved as described previously [4]. Briefly, the distal ureteric dissection was continued beyond the bifurcation of the iliac vessels. Direct contact with the tumour was avoided. Before entering the urinary tract, clips or ligature were placed on the ureter, proximal and distal to the tumour, to prevent tumour spill. For distal ureteric tumours, a wide cuff of bladder was also resected. Indication and extension of lymphadenectomy were left to surgeon's discretion by lack of guidelines during the period of data collection.

We compared the clinical and pathological characteristics of the two surgical technique groups (SU vs RNU) using Student's t-test for continuous variables and the chi-squared test for categorical variables. Univariate and multivariate Cox regression analyses predicting disease recurrence, metastasis and cancer-specific death were performed. Covariates consisted of age, gender, type of surgery (SU or RNU), pT stage, pN stage, tumour grade, LVI and surgical margin status. The actuarial risks of recurrence-free survival (RFS), metastasis-free survival (MFS) and cancer-specific survival (CSS) were calculated using the Kaplan–Meier method and were compared between the two surgical technique groups using the log-rank test. RFS, MFS and CSS were calculated from the date of surgery. A P < 0.05 was considered to indicate statistical significance.

RESULTS

In all, 468 patients were included, 416 patients underwent RNU and 52 underwent SU. The median (interquartile range) follow-up was 26 (10–48) months. In all, 202 patients were excluded because of lack of data concerning pT stage, metastasis status at diagnosis or surgical margin status and 23 were excluded because of metastatic disease at diagnosis. Clinical and pathological characteristics of the two matched study cohorts separately listed for matched and unmatched variables are shown in Table 1. There were no statistically significant difference between the surgical cohorts for the variables used for propensity score matching, including age, gender, grade, pN stage and surgical margins. There were significantly more local advanced tumours (pT3 or pT4) and LVIs in the RNU group (P= 0.001 and P= 0.005, respectively).

Table 1.  Demographic and clinical characteristics
VariableRNUSU P
  1. IQR, interquartile range.

N 41652 
Median (IQR) age, years69.1 (61–76)70.1 (66–76.8)0.18
N (%):   
 Gender:  0.06
  Male291 (70.0)41 (78.8)
  Female125 (30.0)11 (21.2)
 Previous non-muscle invasive bladder cancer114 (27.4)29 (55.7) <0.001
 Pathological stage:   0.001
  pTa/Cis118 (28.4)22 (42.3)
  pT1106 (25.5)12 (23.1)
  pT241 (9.9)12 (23.1)
  pT3131 (31.5)5 (9.6)
  pT420 (4.8)1 (1.9)
 Grade:  0.48
  G123 (5.5)3 (5.8)
  G2144 (34.6)23 (44.2)
  G3249 (59.9)26 (50.0)
 pN stage:  0.38
  pN0124 (29.8)11 (21.2)
  pNx256 (61.5)37 (71.2)
  pN1/236 (8.7)4 (7.7)
 LVI:   0.005
  LVI−335 (80.5)50 (96.2)
  LVI+81 (19.5)2 (3.8)
 Surgical margins:  0.15
  R0378 (90.9)44 (84.6)
  R+38 (9.1)8 (15.4)

The 5-year RFS, MFS and CSS in the SU group were 37%, 81.9% and 87.9%, respectively, vs 47.9%, 85.4% and 86.3% in RNU group; these differences were not statistically significant (P= 0.48, 0.51 and 0.99, respectively; Fig. 1A–C). In univariate analysis, the surgical access did not appear to be a prognostic factor for RFS, MFS or CSS (P= 0.42, 0.53 and 0.94, respectively; Table 2). These results were confirmed in multivariate Cox regression analysis (P= 0.46, 0.54 and 0.56 for RFS, MFS and CSS, respectively).

Figure 1.

Kaplan–Meier curves for: RFS (A), MFS (B), CSS (C) after RNU (red line) or SU (green line) for UUT-UC.

Table 2.  Univariate and multivariable analyses predicting CSS, disease recurrence and metastasis in patients with UUT-UC after SU and RNU
CovariablesCSSRFSMFS
Hazard ratio (95% CI); PHazard ratio (95% CI); PHazard ratio (95% CI); P
Univariable analysisMultivariable analysisUnivariable analysisMultivariable analysisUnivariable analysisMultivariable analysis
Age (continuously coded)1.01 (0.99–1.04); 0.141.01 (0.98–1.01); 0.931.01 (0.99–1.03); 0.23
Gender: male vs female0.98 (0.76–1.26); 0.841.09 (0.81–1.46); 0.591.07 (0.70–1.63); 0.74
Type of surgery SU vs RNU0.99 (0.47–2.08); 0.941.26 (0.58–2.72); 0.561.16 (0.76–1.76); 0.421.17 (0.76–1.80); 0.460.79 (0.40–1.58); 0.531.25 (0.61–2.56); 0.54
Previous non-muscle invasive bladder cancer: yes vs no0.85 (0.19–3.81); 0.83 0.81 (0.44–1.48); 0.501.05 (0.33–3.34); 0.92
Pathological stage:<0.001 0.005 0.02 0.04 <0.001<0.001
 pT1 vs pTa/Tis2.68 (1.07–6.74); 0.031.78 (0.69–4.60); 0.221.62 (1.10–2.38); 0.011.64 (1.11–2.40); 0.011.51 (0.69–3.33); 0.21.34 (0.58–3.10); 0.42
 pT2 vs pTa/Tis2.90 (1.01–8.27); 0.041.44 (0.47–4.36); 0.511.76 (1.13–2.80); 0.011.75 (1.11–2.78); 0.012.60 (1.12–6.04); 0.021.76 (0.69–4.48); 0.21
 pT3 vs pTa/Tis7.40 (3.25–16.85); <0.0013.58 (1.44–8.92); 0.0061.81 (1.24–2.63); 0.0021.63 (1.08–2.46); 0.017.78 (4.13–14.63); <0.0014.30 (1.97–9.35); <0.001
 pT4 vs pTa/Tis18.05 (6.96–46.79); <0.0016.50 (2.13–19.81); 0.0010.74 (0.74–3.32); 0.231.20 (0.76–1.80); 0.466.52 (2.55–16.65); <0.0012.25 (0.74–6.82); 0.15
Grade: 0.42 0.53 0.08 0.56 0.02 0.34
 G2 vs G11.12 (0.31–3.89); 0.830.87 (0.24–3.16); 0.851.61 (0.83–3.13); 0.111.37 (0.70–2.69); 0.330.68 (0.22–2.08); 0.510.52 (0.16–1.67); 0.25
 G3 vs G11.58 (0.89–2.87); 0.071. 39 (0.85–2.92); 0.081.95 (1.02–3.74); 0.041.21 (0.60–2.44); 0.583.44 (1.25–9.44); 0.011.20 (0.38–3.84); 0.71
LVI: yes vs no2.05 (1.19–3.52); 0.0090.59 (0.31–1.11); 0.102.28 (1.03–2.04); <0.0011.16 (0.77–1.75); 0.433.49 (2.37–5.1); <0.0011.38 (0.85–2.23); 0.18
N stage:<0.001 0.007 0.46 <0.001 0.09
 pN1/2 vs pNx2.6 (1.88–3.6); <0.0012.61 (1.78–4.14); 0.0051.32 (0.80–2.19); 0.234.46 (2.67–7.45); <0.0011.65 (0.93–2.92); 0.08
 pN0 vs pNx0.19 (0.05–0.9); 0.960.57 (0.27–1.84); 0.730.90 (0.64–1.28); 0.581.01 (0.58–1.72); 0.980.78 (0.45–1.35); 0.34
Surgical margins: positive vs negative3.03 (1.67–5.50); <0.0011.66 (0.87–3.17); 0.122.28 (1.54–3.37); <0.0012.23 (1.47–3.39); <0.0013.42 (2.11–5.54); <0.0011.98 (1.18–3.31); 0.009

The independent factors predicting RFS were pT stage (P= 0.04) and positive surgical margin (P < 0.001). Concerning MFS, the independent prognostic factors were pT stage (P < 0.001) and positive surgical margin (P= 0.009). The independent factors predicting CSS were pT stage (P= 0.005) and the presence of node metastasis (P= 0.007).

We performed a subgroup analysis concerning unifocal tumours of the distal ureter with a diameter of <2 cm and of low stage (≤T2) from preoperative evaluation (CT and/or MRI). We identified 54 patients treated by RNU and 46 patients by SU (Table 3). There was no statistical difference between the types of surgery. In univariate analysis, type of surgery was not an independent factor for RFS, MFS or CSS (P= 0.96, 0.57 and 0.70, respectively; Table 4). These results were confirmed in multivariate Cox regression analysis (P= 0.94, 0.41 and 0.84 for RFS, MFS and CSS, respectively). Positive surgical margin was the only independent prognostic for RFS and MFS (P= 0.001 and 0.03 respectively).

Table 3.  Demographic and clinical characteristics
VariableRNUSU P
N 5446 
Median (IQR) age, years70.8 (63–76)70.1 (66–76.8)0.61
N (%)   
 Gender:  0.95
  Male42 (77.8)36 (78.8)
  Female12 (22.2)10 (21.7)
 Pathological stage:  0.96
  pTa/Cis27 (50.0)22 (47.8)
  pT114 (25.9)12 (26.1)
  pT213 (24.1)12 (26.1)
 Grade:  0.25
  G19 (16.7)3 (6.5)
  G221 (38.9)22 (47.8)
  G324 (44.4)21 (45.7)
 pN stage:  0.09
  pN08 (14.8)11 (23.9)
  pNx46 (85.2)33 (71.7)
  pN1/20 (0)2 (4.3)
 LVI:  0.62
  LVI−51 (94.4)45 (97.8)
  LVI+3 (5.6)1 (2.2)
 Surgical margins:  0.78
  R049 (90.7)41 (89.1)
  R+5 (9.3)5 (10.9)
Table 4.  Univariate and multivariable analyses predicting CSS, disease recurrence and metastasis after SU and RNU in patients with unifocal UC of the distal ureter with low stage (T2≤) and diameter <2 cm on preoperative evaluation
CovariablesCSSRFSMFS
Hazard ratio (95% CI); PHazard ratio (95% CI); PHazard ratio (95% CI); P
Univariable analysisMultivariable analysisUnivariable analysisMultivariable analysisUnivariable analysisMultivariable analysis
Age (continuously coded)1.04 (0.96–1.13); 0.310.99 (0.96–1.02); 0.941.02 (0.96–1.09); 0.36
Gender: male vs female0.61 (0.07–5.14); 0.650.99 (0.49–2.01); 0.981.27 (0.34–4.72); 0.71
Type of surgery: SU vs RNU1.33 (0.29–5.98); 0.701.18 (0.21–6.42); 0.841.01 (0.56–1.83); 0.961.02 (0.54–1.92); 0.940.72 (0.22–2.28); 0.570.58 (0.16–2.11); 0.41
Pathological stage: 0.36 0.33 0.66 0.27 0.29 0.48
 pT1 vs pTa/Tis1.70 (0.24–12.12); 0.591.44 (0.69–18.95); 0.771.12 (0.58–2.17); 0.721.47 (0.69–3.13); 0.310.96 (0.22–4.03); 0.950.98 (0.16–5.94); 0.99
 pT2 vs pTa/Tis3.59 (0.59–21.65); 0.164.10 (0.58–28.64); 0.150.76 (0.35–1.67); 0.500.68 (0.28–1.64); 0.392.47 (0.70–8.73); 0.022.19 (0.54–8.75); 0.26
Grade: 0.51 0.91 0.66 0.13 0.80 0.85
 G2 vs G11.34 (0.28–4.74); 0.890.74 (0.36–3.04); 0.964.10 (0.94–17.79); 0.053.93 (0.89–17.32); 0.070.74 (0.45–1.06); 0.930.87 (0.32–1.27); 0.94
 G3 vs G11.52 (0.76–2.74); 0.121. 65 (0.93–4.72); 0.973.61 (0.84–15.49); 0.082.57 (0.56–11.79); 0.222.32 (0.65–8.83); 0.821.31 (0.56–3.72); 0.94
Lymphovascular invasion: yes vs no0.82 (0.19–3.52); 0.781.30 (0.31–5.40); 0.712.35 (0.30–18.40); 0.41
N stage: 0.22 0.17 0.14
 pN1/2 vs pNx6.56 (0.74–58.17); 0.090.78 (0.10–5.73); 0.815.87 (0.74–46.46); 0.09
 pN0 vs pNx0.90 (0.10–7.73); 0.920.40 (0.15–1.05); 0.060.40 (0.05–3.15); 0.38
Surgical margins: positive vs negative3.12 (1.40–6.97); 0.0052.91 (0.95–8.90); 0.061.95 (1.31–2.91); 0.0012.28 (1.39–3.73); 0.0012.45 (1.32–4.55); 0.0042.11 (1.07–5.30); 0.03

DISCUSSION

Because of the aggressive pattern of UC, RNU with bladder cuff removal remains the ‘gold-standard’ treatment of UUT lesions [1,2]. Under certain conditions, renal-sparing surgery for UUT-UC could be an alternative radical treatment. The conservative approach, which is the first choice to treat patients for whom radical surgery is contraindicated, could be considered for select patients with normal contralateral kidney and low-risk UUT-UC [1].

Considering recent developments in imaging techniques and predictive tools used to better stratify patients according to risk category, these can also play important roles in selecting patients for conservative management. In fact, the developments of imaging (CT and MRI) can now better localise and characterise UUT-UC preoperatively (stage p > T2, size) [5,6]. Additionally, determining tumour grade after ureteroscopic biopsy is an increasingly common diagnostic tool to predict final pathological stage [7,8]. Recently, a multicentre study has reported the combination of three variables (presence of hydronephrosis, urinary cytology and biopsy grade) as an accurate predictive tool of the final pathological stage [9]. Therefore, the selection of candidates for conservative treatment seems to be safer today than ever.

Otherwise, conservative procedures allow the morbidity associated with RNU to be reduced. Moreover, there is a real benefit in terms of cost in choosing renal preservation treatment in cases of necessity indication, as emphasised by a recent cost-effectiveness study [10]. Moreover, an elective resection of the distal ureter with re-implantation could be an interesting option for patients with borderline renal function and/or when adjuvant chemotherapy is being considered.

During the last 15 years, endoscopic management of UUT-UC has expanded considerably in both the technical advances in optics and energy sources (flexible ureteroscope, holmium: yttrium-aluminium-garnet [YAG] laser) and has changed the face of conservative treatment [11]. However, one drawback of endoscopic approaches is the paucity of specimens for pathological evaluation where devices such as the laser were used to treat the tumour, in opposition with SU [12]. Another obvious drawback is the cost that such complex endoscopic devices require for the institutions that decide to buy them.

On the other hand, SU remains a good treatment option, still validated by international guidelines in the case of localised and small lower ureteric tumours [1]. However, invasive SU, i.e. requiring the opening of the urinary tract, offers the possibility of complete resection of visible lesions with extemporaneous analysis of excision margins. Moreover, even if the oncological interest and extension are still debated, lymphadenectomy can be performed during SU to better define the tumour staging than an endoscopic conservative treatment can. The open procedure is the standard for SU, surgical feasibility and early oncological outcomes of laparoscopic distal ureterectomy in patients with low-grade UUT-UC have been reported [1].

Most series describing oncological results after SU are retrospective, with few patients and heterogeneous samples. In 1972, Mazeman [13] showed that the incidence of local recurrence was similar whether the patient undergoes RNU or distal ureterectomy with re-implantation. These results were verified by others, and the rate of recurrence was estimated at between 10 and 25% [14,15]. This is especially true for low-stage and low-grade tumours.

In patients with a tumour grade 1 (WHO 1973), Murphy et al.[16] reported a 5-year overall survival of 77% in the case of SU vs 88% in RNU. For grade 2 tumours, conservative treatment had a weaker survival (46% vs 90%), and for grade 3 tumours, the prognosis was worse regardless of the treatment procedure. Zungri et al.[15] reported that 5-year overall survival was similar in SU and RNU in cases of stage pTa tumours. This rate was 70% in the conservative group against 96% in the radical group for pT1 tumours; invasive stages had a poor prognosis in both groups.

Recently, a retrospective multicentre study involving >2000 patients showed no significant difference in survival between patients specifically treated with RNU with bladder cuff excision compared with those treated with SU [17,18]. Their 5-year CSS rates for SU and RNU with bladder cuff excision were 86.6% and 82.2%, respectively (P≥ 0.05). After stratification according to pT stage (pT1–2 vs pT3–4), in patients with pT1–2 ureteric UUT-UC, the 5-year CSS rates were 91.2% for the SU group vs 88.9% for the RNU group, respectively (P > 0.4). For patients with locally advanced (pT3–4) ureteric UUT-UC, theses survival rates were 71.9% and 67.8%, respectively (P > 0.2). In the present series, results for CSS were comparable to previous data, with 87.9% of patients treated by SU alive at 5 years vs 86.3% in case of radical surgery (P= 0.99). Furthermore, the sub-group analysis concerning only distal and unifocal ureteric locations did not show a significant effect of surgery type.

One limitation of the previous international study was to report only CSS [17]. While the risk of recurrence after conservative treatment is an important issue requiring enhanced surveillance, no analysis of RFS or MFS was reported. To our knowledge, the present results concerning RFS after SU compared with RNU concern the largest sample of any such published report. Therefore, after relative standardisation of follow-up to French national guidelines, the present results suggest no statically significant difference in urothelial recurrence frequency between the SU and RNU groups (P= 0.48). This absence of significant difference was also found for MFS. As the role of ureteric location is discussed in terms of recurrence and the occurrence of ureteric metastasis for UUT-UC, we performed a subgroup analysis for this group of patients [19,20]. There was no significant difference between patients treated with SU and RNU for RFS or MFS (P= 0.94 and 0.41, respectively).

Another interesting finding of the present study was the prognostic impact of positive surgical margins. Recently, Novara et al.[21] showed that positive surgical margin after radical cystectomy was an independent prognostic factor for CSS and RFS. Concerning UUT-UC, some authors have described the prognostic value of positive surgical margins for overall survival and CSS in univariate analyses [22,23]. For infiltrative disease (≥pT2), Ku et al.[22] have confirmed positive surgical margins as an independent prognostic factor for bladder RFS. Additionally, careful analyses by pathologists of UUT-UC excision margins and reports on their positivity for residual cancer cells are necessary. Positive surgical margins should be considered during follow-up and also for adjuvant chemotherapy administration.

The present study has some limitations due to its retrospective design. To avoid bias, we carried out a subgroup analysis using propensity score matching. The criteria (unifocal tumour of the distal ureter, <2 cm in diameter, without evidence of extra-ureteric extension on preoperative imaging) were those used to select patients eligible for US as conservative treatment without any imperative indication. The two subgroups were comparable for age, gender, pT stage, grade, pN stage, LVI and surgical margin status (Table 3). Regarding the type of surgery, there was also no significant difference in survival (Table 4).

Another criticism of the present study could be the lack of standardisation of the selection criteria and surgical technique in relation to the number of participating centres. However, all centres included in the study followed the recommendations of our national society of urology [24–26].

Additionally, as the inclusion period of the present study began in 1995, LVI was not systematically assessed and/or reported by pathologists. The lack of information concerning LVI constitutes a bias for the interpretation of its prognostic significance and statistical comparisons between the SU and RNU groups. In addition, patients in the present series underwent surgical treatment for UUT-UC by multiple surgeons, and multiple pathologists performed the histopathological analyses. The lack of a centralised pathological review is balanced by relative standardisation of practices imposed by our national urological and pathological societies' guidelines.

Finally, regardless of size, all currently published retrospective studies provide only level 3 evidence and thus show only the tendencies that reflect current practices and feelings towards UUT-UC. A prospective database that would assess this question is urgently needed to reach a conclusion for this on-going debate.

In conclusion, short-term oncological data obtained after conservative management of UUT-UC with SU were comparable to RNU for the management of UUT-UC in select cases. Thus, this strategy should be considered in certain cases where the tumour is unique and of limited size. In all other cases, RNU remains the gold-standard treatment.

ACKNOWLEDGMENTS

The authors would like to kindly thank other collaborators from the French collaborative national database on UUT-UC: François Audenet, Alexis A. Berod, Jean A. Long, Alain Ruffion, Emilie Adam, Olivier Cussenot, Jean N. Cornu, Alain Houlgatte, Véronique Phé, François X. Nouhaud, Henri Bensadoun, Antoine Valéri, Franck Delage, Mathieu Roumiguié, Thomas Polguer, Nicolas Hoareau, Evanguelos Xylinas, Charlotte Maurin, Marie-Dominique Azemar, Frédéric Staerman, Luc Cormier, Stéphane Droupy, Pascal Gres, Gilles Karsenty, Aurélien Descazeaud, Solène Gardic, Nicolas Brichart, Franck Bruyère.

CONFLICT OF INTEREST

None declared.

Ancillary