Oncologic Outcomes of Patients Undergoing Diagnostic Ureteroscopy Before Radical Nephroureterectomy for Upper Urinary Tract Urothelial Carcinomas: A Systematic Review and Meta-Analysis

Abstract

Background:

Owing to the aggressive nature of radical nephroureterectomy (RNU), it is extremely important for the diagnostic modalities to be accurate. The European Association of Urology guidelines recommend that diagnostic ureteroscopy (URS) and biopsy be performed for upper urinary tract urothelial carcinomas (UUTCs). But that diagnostic URS adversely affects oncologic outcomes still exist. In the current systematic review and meta-analysis, we had attempted to evaluate oncologic outcome of diagnostic URS before RNU.

Methods:

All relevant articles were retrieved from PubMed, Embase, Web of Science, and the Cochrane Library. Endpoint events were recurrence-free survival, metastasis-free survival, cancer-specific survival, and overall survival.

Results:

Compared with patients who underwent RNU alone, those who underwent diagnostic URS before RNU had significantly higher intravesical recurrence rate (HR = 1.42, 95% CI: 1.24–1.61, I² = 37%). Overall, no significant difference was found in CSS (HR = 0.72, 95% CI: 0.51–1.03, I² = 0%) and OS (HR = 0.73, 95% CI: 0.45–1.19, I² = 0%) between patients who underwent diagnostic URS and those who did not. Meanwhile, there was lack of evidence indicating that the risk of tumor metastasis increased after diagnostic URS (HR = 0.97, 95% CI: 0.74–1.26, I² = 0%).

Conclusions:

Diagnostic URS before RNU does not seem to compromise long-term survival outcomes, even though it is associated with a higher rate of intravesical recurrence (IVR). Our findings suggest that further investigation, especially through prospective studies, should focus on decreasing the rate of IVR by administration of intravesical chemotherapy immediately after diagnostic URS.

Background

Upper urinary tract urothelial carcinomas (UUTCs) are quite rare and account for 5%–10% of all urothelial carcinomas.¹ Radical nephroureterectomy (RNU) has been the standard of treatment for UTUCs, although kidney-sparing surgery and endoscopic treatment might be alternatives for select patients.² Owing to the aggressive nature of RNU, it is extremely important for the diagnostic modalities to be accurate. Although the specificity of urine cytology is reportedly 99%, its sensitivity is highly dependent on tumor grade and stage, ranging from 38% to 51%.³ The European Association of Urology guidelines (EAU) state that computed tomographic urography (CTU) is the preferred imaging technique for patients with UTUCs, as it is more sensitive and specific than magnetic resonance imaging-based urography.² However, the lesion diameter greatly affects the sensitivity of CTU and it has been found to be 89% for lesions of diameter less than 5 mm, but decreases to 40% for lesions of diameter less than 3 mm.³ Previous studies reported that 2.9%–12.8% of all radiologically suspected masses in the upper urinary tract are proven to be benign morphological changes on RNU,^4–6 including inflammatory pseudotumors, benign urothelial cysts, and capillary hemangiomas. Therefore, pathological diagnosis is essential before surgery to avoid unnecessary RNU.

Ureteroscopy (URS) is a useful tool for visual detection and to provide a working channel during UTUC biopsy. In their prospective study, Grahn et al.⁷ reported that diagnostic URS has significantly higher specificity and accuracy than multiphase CTU. Golan et al.⁸ and Sadow et al.⁹ reported that all suspicious filing defects of size less than 5 mm discovered on CTU were found to be benign morphological changes on diagnostic URS. In addition, some experts reported that the rate of the specificity of diagnostic URS with biopsy reaches 100%.^10,11 Nonetheless, concerns related to the increased risk of recurrence, progression, and metastases following diagnostic URS persist. First, the surge pressure in the renal pelvis may increase the possibility of dissemination through submucosal vessels or lymphatic tubes during the irrigation step of diagnostic URS, and this would counteract the survival benefits.^12,13 Second, slough tumor cells along with the irrigation fluid can flow back into the bladder and get implanted in the mucous membrane, which increases the risk of intravesical recurrence.

Previous studies have evaluated the oncologic outcomes of diagnostic URS using different endpoint events, such as recurrence-free survival, intravesical recurrence (IVR), metastasis-free survival (MFS), cancer-specific survival (CSS), and overall survival (OS). For this reason, it is difficult to draw constructive conclusions without a systematic analysis of current studies. The aim of the present systematic review and meta-analysis was to compare and evaluate oncologic outcomes of patients with and without diagnostic URS before RNU.

Methods

Identification of studies

All relevant articles were retrieved from PubMed, Embase, Web of Science, and the Cochrane Library using the following search terms: (“upper tract urothelial carcinomas” OR “ureteral carcinoma” OR “pelvic carcinomas”) AND (“Ureteroscopy” OR “Ureteroscopic surgery”) AND (“intravesical recurrence” OR “cancer specific survival” OR “overall survival” OR “metastases free survival”); last updated: 10 May, 2017).

Inclusion and exclusion criteria

Study eligibility was defined on the basis of the participants, (P), intervention (I), comparator (C), and outcome (O) (PICO) approach, as recommended in the Systematic Reviews and Meta-analysis (PRISMA) guidelines ¹⁴. In the present analysis, a study was included if it met the following criteria: it included patients with UTUCs (P) who did (I) or did not (C) undergo diagnostic URS following RNU, and analyzed oncologic survival outcomes (O). Additionally, the studies needed to (i) compare oncologic outcomes between patients who underwent diagnostic URS following RNU and those who did not undergo diagnostic URS following RNU; (ii) present results included at least one of the following outcomes: IVR, CSS, OS, and MFS; and (iii) be published in English. Studies were excluded if they were editorials, letters to the editor, reviews, case reports, or meeting abstracts. The primary results of the literature search were exported to EndNote, and 711 duplications were identified.

Data extraction and outcomes of interest

Data from the included reports were extracted independently by two investigators (Z.F.L. and S.Y. Z.). All disagreements were resolved by adjudication by the senior author (Z.W.L.). The extracted data included the name of the first author, year of publication, study design, number of participants in each treatment group, participants' age and gender, and the outcomes of interest. Our primary outcome was IVR, and the secondary outcomes were CSS, MFS, and OS.

Quality assessment

Criteria established by the Center for Evidence-Based Medicine in Oxford, UK, were used to rate the level of evidence in the included studies ¹⁵. The modified Newcastle–Ottawa scale, a validated tool to evaluate nonrandomized studies, was used to assess the risk of bias in the nonrandomized control trials. This scale involves three factors, namely, patient selection, comparability of the study groups, and assessment of outcomes. We determined that the included studies were high-quality articles with scores exceeding 7.

Statistical analysis

A meta-analysis was conducted using Review Manager 5.0 (Cochrane Collaboration, Oxford, United Kingdom) to examine studies comparing the same pair of treatments. Generally, the hazards ratio (HR) is considered the only summary statistic for censored outcomes as described by Parmar¹⁶. In the present study, we used the HRs and 95% confidence intervals (CIs) to determine the relationship between URS and oncologic survival outcomes (IVR, CSS, MFS, and OS). Standard deviations were calculated using the technique described by Hozo et al.¹⁷ for continuous variables, which were presented as mean and range values in the included studies. Statistical heterogeneity among the reports was assessed using the chi-square test with significance set at P < .05, the I² statistic, and a forest plot. The random-effects model was used when data were significantly heterogeneous (I² > 50%); otherwise, the fixed-effects model was used. Subgroup analyses for IVR were conducted to evaluate the influence of diagnostic URS with biopsy and history of bladder cancer. Sensitivity analyses were performed to explore the potential sources of heterogeneity and eliminating the effect of low-quality articles. Potential publication bias was screened using funnel plots.

Results

Overall study characteristics and eligibility

The flow of study selection and the funnel plot of all included studies were presented as Figures 1 and 2, respectively. Overall, after 1092 studies were excluded, 9 studies met the inclusion criteria and were included. Data of patients from the 9 included studies^18–26 were examined in this systematic review and meta-analysis. In total, 3198 patients were analyzed, including 1038 patients who underwent diagnostic URS before RNU and 2160 patients who did not. None of the 3198 patients underwent kidney-sparing surgery or endoscopic treatment. With the exception of one study published in 1999, the remaining studies were published after 2000, including one published in 2010 and seven published in the last 5 years. All included studies were retrospective. Four of the included studies were considered to be of high quality (7 scores) and the remaining five had scores of 5–6 (Table 1).

FIG. 1.

Study selection flow.

FIG. 2.

Funnel plot of all included studies. IVR was compared between patients who underwent diagnostic URS before RNU and those who did not. SE, standard error; OR, odds ratio; IVR, intravesical recurrence; RNU, radical nephroureterectomy; URS, ureteroscopy.

Table 1.

Assessment of the Quality of Included Studies

					No. of patients		Age		Gender, male		Follow-up time, months
Publications	Country	Study design	Level of evidence	Matching	URS+	URS−	URS+	URS−	URS+	URS−	URS+	URS−	Quality score
Sung et al.²³	Korea	R	3b	1,3,4,5,6	282	348	64	65	212	253	30.1	39.3	★★★★★★★
Luo et al.²⁶	Taiwan	R	3b	1,2,3,4,5,6	115	281	65.9	66.6	54	136	42.4	38.9	★★★★★★★
Nison et al.²⁰	France	R	3b	3	170	342	68.8	70.1	112	236	21.4	24	★★★★★★★
Liu et al.¹⁹	China	R	3b	1,3,6	81	583	65.9	66.6	31	264	NA	NA	★★★★★★
Sankin et al.²²	USA	R	3b	1,2	144	57	70	71	87	23	NA	NA	★★★★★★
Yoo et al.²¹	Korea	R	3b	1,2,3,4,5,6	69	318	63.8	63.9	52	232	NA	NA	★★★★★★★
Ishikawa et al.²⁴	Japan	R	3b	NA	55	153	NA	NA	36	103	NA	NA	★★★★★
Hendin et al.²⁵	USA	R	3b	1,2	48	48	63.2	67.5	37	31	50.4	42.4	★★★★★★
Lee et al.¹⁸	Korea	R	3b	1,2,6	74	30	65.7	67.5	48	23	29.5	37	★★★★★★

Matching: 1, tumor grade; 2, tumor stage; 3, cis; 4, lymphovascular invasion; 5, history of bladder tumor; 6, multiple tumors.

R, retrospective; NA, data not available; URS, ureteroscopy.

Patients' demographic and clinicopathological features

Diagnostic URS was far more likely to be selected for patients with smaller, low-grade, <T2 stage, ureter tumors (Table 2). This choice was not affected by American Society of Anesthesiologists score (Table 3).

Table 2.

Pathological Characteristics in the Included Studies

	Tumor size, median (IQR)			Tumor grade (%)			Tumor stage (%)			LN status (%)
Publications	URS+	URS−	P value	URS+	URS−	P value	URS+	URS−	P value	URS+	URS−	P value
Sung et al.²³	3.0 (1.8–4.0)	4.0 (3.0–5.5)	<0.01	G3: 124 (45.4)	153 (44.5)	.82	Ta:59 (20.9)	50 (14.4)	.01	LN+:20 (7.1)	36 (10.3)	.15
							T1:84 (29.8)	69 (19.8)
							T2:48 (17.0)	55 (15.8)
							T3/T4:91 (32.3)	174 (50.0)
Luo et al.²⁶	NA	NA	—	High:102 (88.7)	255 (90.7)	.53	≤T2:90 (78.3)	199 (70.8)	.13	NA	NA	—
				Low:13 (11.3)	26 (9.3)		>T2:25 (21.7)	82 (29.2)
Nison et al.²⁰	NA	NA	—	G1:28 (16.5)	34 (9.9)	.03	Ta/cis:63 (37.1)	78 (22.8)	<.01	LNx:113 (66.5)	213 (62.3)	.10
				G2:56 (32.9)	98 (28.7)		T1:37 (21.8)	74 (21.6)		LN0:50 (29.4)	97 (28.4)
				G3:86 (50.6)	210 (61.4)		T2:12 (7.1)	41 (12.0)		LN+:7 (4.1)	32 (9.4)
							T3:52 (30.6)	124 (36.3)
							T4:6 (3.6)	25 (7.3)
Liu et al.¹⁹	≥3 cm:28 (34.6%)	335 (57.5%)	<0.01	High:30 (37.0)	253 (43.4)	.28	≤T2:65 (80.2)	393 (67.4)	.02	LN0:78 (96.3)	539 (92.5)	.21
	<3 cm: 53 (65.4%)	248 (42.5%)		Low:51 (63.0)	330 (56.6)		>T2:16 (19.8)	190 (32.6)		LN+:3 (3.7)	44 (7.5)
Sankin et al.²²	3.1 (2.0–4.5)	3.5 (2.5–4.5)	.15	High:112 (78)	49 (86)	.20	Tis:3 (2.1)	0 (0)	.03	NA	NA	—
							Ta:49 (34)	11 (19)
							T1:27 (19)	9 (16)
							T2:26 (18)	16 (28)
							T3:37 (26)	16 (28)
							T4:2 (1.4)	5 (8.8)
Yoo et al.²¹	NA	NA	—	High:32 (47.1)	158 (49.8)	.69	Ta/T1:37 (53.6)	170 (53.5)	.51	LN+:3 (3.4)	23 (7.20)	.60
				Low:36 (52.9)	159 (50.2)		T2:15 (21.8)	53 (16.7)
							T3/T4:17 (24.6)	95 (29.8)
Ishikawa et al.²⁴	NA	NA	—	High:12 (21.8)	54 (35.3)	.06	Ta–is:13 (23.6)	28 (18.3)	.16	LN0:27 (49.1)	77 (50.3)	.13
				Low:43 (78.2)	99 (64.7)		T1:17 (30.9)	42 (27.5)		LNx:27 (49.1)	63 (41.2)
							T2:13 (23.6)	31 (20.3)		LN+:1 (1.8)	13 (8.5)
							T3:12 (21.8)	44 (28.8)
							T4:0 (0)	8 (5.2)
Hendin et al.²⁵	NA	NA	—	G1:9 (20)	6 (12)	.09	Tis:4 (8)	2 (4)	.44	NA	NA	—
				G2:23 (51)	19 (40)		Ta:24 (50)	22 (46)	.84
				G3:13 (29)	23 (48)		T1:9 (19)	7 (15)	.79
							T2:1 (2)	2 (4)	.62
							T3:9 (19)	13 (27)	.34
							T4:0 (0)	1 (2)	.50
Lee et al.¹⁸	NA	NA	—	High:56 (75.7)	23 (76.7)	.86	T < 1:31 (41.9)	9 (30.0)	.53	NA	NA	—
				Low:18 (24.3)	7 (23.3)		T = 2:33 (44.6)	14 (46.7)
							T = 3:10 (13.5)	7 (23.3)

NA, data not applicable.

Table 3.

Clinical Characteristics of Included Studies

	No. of patients		ASA scores, %			No. of multiple tumor(%)			History of bladder tumor(%)
Publications	URS+	URS−	URS+	URS−	P value	URS+	URS−	P value	URS+	URS−	P value
Sung et al.²³	282	348	≥II: 69.9	67.5	.531	81 (28.7)	87 (25.0)	0.293	44 (15.6)	79 (22.7)	.025
Luo et al.²⁶	115	281	NA	NA		38 (33.0)	92 (32.7)	0.953	31 (27)	77 (27.4)	.928
Nison et al.²⁰	170	342	I:21.38	20.46	.23	NA	NA	—	NA	NA	NA
			II:58.49	51.16
			III:18.87	27.39
			IV:1.26	0.99
Liu et al.¹⁹	81	583	NA	NA	—	27 (33.3)	132 (22.6)	0.035	0	0	—
Sankin et al.²²	144	57	NA	NA	—	NA	NA	—	NA	NA	—
Yoo et al.²¹	69	318	NA	NA	—	14 (20.3)	55 (17.3)	0.603	13 (18.8)	45 (14.2)	.352
Ishikawa et al.²⁴	55	153	NA	NA	—	7 (12.7)	41 (26.8)	0.0263	13 (23.6)	127 (17.0)	.288
Hendin et al.²⁵	48	48	NA	NA	—	NA	NA	—	0	0	—
Lee et al.¹⁸	74	30	NA	NA	—	11 (14.9)	10 (33.3)	0.091	0	0	—

	No. of patients		Surgical approach (Open/Laparo)			Bladder cuff excision (%)			Tumor location (%)
Publications	URS+	URS−	URS+	URS−	P value	URS+	URS−	P value	URS+	URS−	P value
Sung et al.²³	282	348	140/142	252/96	<0.001	All	All	—	NA	NA
Luo et al.²⁶	115	281	NA	NA	—	All	All	—	Renal pelvis:62 (53.9)	203 (72.2)	<0.001
									Ureter:83 (72.2)	149 (53.0)
Nison et al.²⁰	170	342	NA	NA	—	NA	NA	—	Pelvis:80 (47.06)	197 (57.60)	.02
									Ureter:71 (41.76)	101 (29.53)
									Multifocal:19 (11.18)	44 (12.87)
Liu et al.¹⁹	81	583	58/23	387/196	.349	All	All	—	Pelvis:26 (32.1)	342 (58.7)	<0.001
									Ureter:55 (67.9)	241 (41.3)
Sankin et al.²²	144	57	NA	NA	—	NA	NA	—	Plevis:83 (58)	44 (77)	.028
									Ureter:40 (28)	7 (12)
									Both pelvis and ureter:21 (15)	6 (11)
Yoo et al.²¹	69	318	30/39	152/166	.595	All	All	—	Pelvis:27 (39.1)	158 (49.7)	.143
									Ureter:42 (60.9)	160 (50.3)
Ishikawa et al.²⁴	55	153	42/13	123/30	.5312	All	All	—	Pelvis:27 (49.0)	84 (54.9)	.3858
									Upper ureter:15 (27.3)	28 (18.3)
									Lower ureter:13 (23.6)	41 (26.8)
Hendin et al.²⁵	48	48	NA	NA	—	All	All	—	Ureter:23 (47.9)	14 (29.2)	.09
Lee et al.¹⁸	74	30	42/32	14/16	.609	NA	NA	—	Pelvic and ureter:68 (91.9)	21 (70)	.011
									Ureter:6 (8.1)	9 (30.0)

NA, data not available; URS, ureteroscopy; ASA, American Society of Anesthesiologists.

Intravesical recurrence

We found that compared with patients who underwent RNU alone, those who underwent diagnostic URS before RNU had a significantly higher IVR (HR = 1.42, 95% CI: 1.24–1.61, I² = 37%, Fig. 3A). Nison et al.²⁰ defined disease recurrence as bladder or contralateral upper urinary tract recurrence. After the study by Nison et al.²⁰ was excluded, the heterogeneity decreased from 37% to 9%, but diagnostic URS still significantly increased the rate of IVR (HR = 1.50, 95% CI: 1.31–1.73, I² = 9%, Fig. 3B). After exclusively including patients with RNU and bladder cuff excision, diagnostic URS associated with high risk of IVR (HR = 1.46, 95% CI: 1.26–1.69, I² = 0%, Fig. 3C). We noticed that five literatures had no neoadjuvant or adjuvant chemotherapy and their outcome shows that diagnostic URS was a risk factor for IVR (HR = 1.34, 95% CI: 1.13–1.59, I² = 0%; Fig. 3D). Subgroup analysis, including patients who did not receive intravesical chemotherapy after RNU (HR = 1.50, 95% CI: 1.25–1.78, I² = 0%, Fig. 3E), and excluding patients with a history of bladder cancer (HR = 1.71, 95% CI: 1.38–2.11, I² = 0%; Fig. 4A), showed the same result. The biopsy during the diagnostic URS did not significantly influence the rate of IVR (HR = 0.81, 95% CI: 0.61–1.07, I² = 0%, Fig. 4B). Lee et al.¹⁸ and Sung et al.²³ classified patients into two groups according to the period between diagnostic URS and RNU: one group underwent diagnostic URS and RNU on the same day, whereas RNU was delayed in the other group. Therefore, subgroup analysis between patients who received concurrent RNU after diagnostic URS and those who received delayed RNU was presented in Figure 4C. Totally, there was no evidence that delayed RNU significantly increased the risk of IVR (HR = 0.90, 95% CI: 0.61–1.32, I² = 1%, Fig. 4C).

FIG. 3.

Forest plot and meta-analysis of IVR. (A) Overall comparison of IVR between patients who underwent diagnostic URS before RNU and those who did not. (B) Sensitivity analysis of IVR for diagnostic URS. (C) Subgroup analysis of IVR in patients with RNU and bladder cuff excision. (D) Subgroup analysis of IVR in patients without neoadjuvant or adjuvant chemotherapy. (E) Subgroup analysis of IVR in patients who did not receive intravesical chemotherapy after RNU.

FIG. 4.

Forest plot and meta-analysis of IVR. (A) Subgroup analysis of IVR in patients without a history of bladder tumors. (B) Subgroup analysis of IVR between patients who underwent diagnostic URS with biopsy and those who underwent diagnostic URS alone. (C) Subgroup analysis of IVR between patients who received concurrent RNU after diagnostic URS and those who received delayed RNU.

Long-term survival endpoint events (MFS, CSS, and OS)

The heterogeneity in the subgroup analyses for long-term survival endpoint events were mostly low (I² = 0%). Overall, no significant difference was found in CSS (HR = 0.72, 95% CI: 0.51–1.03, I² = 0%, Fig. 5A) and OS (HR = 0.73, 95% CI: 0.45–1.19, I² = 0%, Fig. 5B) between patients who underwent diagnostic URS and those who did not. Meanwhile, there was lack of evidence indicating that the risk of tumor metastasis increased after diagnostic URS (HR = 0.97, 95% CI: 0.74–1.26, I² = 0%, Fig. 5C).

FIG. 5.

Forest plot and meta-analysis of long-term oncologic survival outcomes. Outcomes were compared between patients who underwent diagnostic URS before RNU and those who did not. (A) Cancer-specific survival. (B) Overall survival. (C) Metastasis-free survival.

Discussion

RNU is recommended as standard surgery for patients with UUTCs. However, it is an extremely aggressive surgery and is generally accompanied by loss of half the renal unit. Therefore, accurate diagnosis plays an essential role in the treatment of patients. In fact, diagnostic URS has a higher sensitivity and specificity than other imaging modalities,^7,8 but the grade for which the EAU guidelines recommend the use of this technique is C.² Currently, concerns about the increased risk of recurrence, progression, and metastasis following diagnostic URS persist.^12,27 Minowada et al.²⁸ reported that three (75%) patients with grade three UUTCs who underwent diagnostic URS before RNU developed systematic metastases within a year of the surgery.

We found that diagnostic URS before RNU was associated with significantly higher rate of IVR, but not poorer MFS, CSS, or OS. An explanation for the increased risk of IVR after diagnostic URS is that tumor cells get detached and are shed following the fluid irrigation step of URS. However, heterogeneity across the nine studies included for evaluation of the IVR was modest (I² = 37%). According to some studies, patient characteristics (gender and history of preoperative chronic kidney disease), type of surgery (e.g., laparoscopic or open), tumor differentiation (T stage, number, etc.), and a history of bladder cancer also affect the IVR rate.^29–32 Therefore, we conducted some subgroup analyses with the aim of reducing heterogeneity.

Because patients with a history of bladder urothelial carcinomas have a higher possibility of IVR after RNU than those without these conditions,^29–32 we performed a subgroup analysis of patients without a history of bladder urothelial carcinomas. Similar to previous studies, we found that diagnostic URS was associated with a significant higher rate of IVR in these patients (I² = 0%, Fig. 4A). Subsequently, we analyzed that literatures included patients with RNU and bladder cuff excision and the outcome show diagnostic URS also significantly increased the risk of IVR (I² = 0%, Fig. 3C). Recently, Nissar et al.³³ reported that diagnostic URS with biopsy before RNU were significantly associated with a higher risk of IVR compared with diagnostic URS without biopsy before RNU. However, we found that diagnostic URS with biopsy did not significantly increase the risk of IVR (I² = 0%, Fig. 4B). Since it is known that prophylactic intravesical chemotherapy could significantly decrease the IVR ³⁴, we performed a subgroup analysis excluding patients receiving intravesical chemotherapy following RNU, and our findings confirmed that diagnostic URS appeared to increase the risk of IVR (I² = 0%, Fig. 3E). Some experts are also concerned that a prolonged interval between diagnostic URS and RNU might increase the risk of tumor seeding. The findings of our study did not show that concurrent RNU was beneficial toward decreasing the risk of IVR (I² = 1%, Fig. 4C). At present, whether or not IVR after RNU affects the survival benefits remains debatable. Yamashita et al.³⁵ reported that IVR was associated with the risk of poor CSS (HR = 4.27, P = .016), whereas Xylinas et al.³¹ did not observe such a relationship between IVR and CSS (log rank P = .06).

The high irrigation pressure in the renal pelvis and ureter caused by diagnostic URS may contribute to tumor metastasis through the vascular or lymphatic tube. Our study found similar long-term survival (MFS, CSS, and OS) among patients who underwent diagnostic URS before RNU to those who underwent RNU alone. Notably, we found that patients who underwent diagnostic URS had been conferred potential tendency for better CSS and OS. Indeed, we were not able to conclusively interpret this phenomenon because we were unable to acquire individual data. Keeley et al.³⁶ noted that diagnostic URS combined with biopsy could provide precise information about tumor stage and grade. We speculate that accurate diagnosis of tumor stage and grade contributes to the decision about the necessity of lymph node dissection during RNU, which shows survival benefits in a more extended scope dissection ^{37, 38}. From the pathological viewpoint, the selection bias in patients from both groups is responsible for that potential tendency toward better CSS and OS. For example, patients who have not undergone diagnostic URS usually have predisposing factors, such as a higher tumor grade and T stage and larger tumors than those who have not (Table 2).

Not only did the URS allow direct visual examination of radiologic filling defects, but it also provided a working channel for the biopsy and other advanced imaging techniques. The prognostic information provided by imaging as well as pathological examination is critical for urologists to categorize patients so they receive the optimal treatment. Seisen et al.³⁹ established a risk-adapted selection system for patients with UUTCs who would benefit more from conservative treatment, in which diagnostic URS plays an essential role. The main strength of our study is that application of diagnostic URS before RNU does not compromise long-term survival outcomes. Furthermore, our findings mitigate concerns related to the poorer survival benefits in patients who undergo diagnostic URS before kidney-sparing surgery or endoscopic management.

The findings of the present study should be considered taking into consideration the limitations of the retrospective design of all included studies. First, the differences in clinicopathological features between groups may have introduced some bias in the results. Moreover, it is not possible to divide the patient groups between those with lower stage and grade tumors (T1-2 or low grade) and higher stage and grade tumors (T3-4 or high grade), and then further investigate the influence of URS on those subgroups. Future studies should mostly focus on the prognostic information provided by diagnostic URS for low-risk patients and high-risk patients separately.

In conclusion, accurate preoperative diagnosis and classification of UUTCs is critical for decision making regarding treatment. Diagnostic URS before RNU does not seem to compromise long-term survival outcomes, even though it is associated with a higher rate of IVR. Our findings suggest that further investigation, especially through prospective studies, should focus on decreasing the rate of IVR by administration of intravesical chemotherapy immediately after diagnostic URS.

Footnotes

Acknowledgments

The meta-analysis was designed by L.Z.F., Z.S.Y., and Z.F.J., and L.Z.W., L.Z.F., and Z.S.Y. performed the systemic review and data summary. Z.S.Y., L.X.D., G.S.J., and W.Y.J. were responsible for the analysis of pooled data. The article was written and modified by L.Z.F., Z.F.J., and L.Z.W. All authors read and approved the final article.

Disclosure Statement

No competing financial interests exist.

References

Green

, Rink

, Xylinas

, et al. Urothelial carcinoma of the bladder and the upper tract: Disparate twins. J Urol, 2013; 189:1214–1221.

Roupret

, Babjuk

, Comperat

, et al. European association of urology guidelines on upper urinary tract urothelial cell carcinoma: 2015 Update. Eur Urol, 2015; 68:868–879.

Baard

, de Bruin

, Zondervan

, Kamphuis

, de la Rosette

, Laguna

. Diagnostic dilemmas in patients with upper tract urothelial carcinoma. Nat Rev Urol, 2017; 14:181–191.

Chitale

, Mbakada

, Irving

, Burgess

. Nephroureterectomy for transitional cell carcinoma - the value of pre-operative histology. Ann R Coll Surg Engl, 2008; 90:45–50.

Hong

, Kwon

, You

, et al. Incidence of benign results after laparoscopic radical nephroureterectomy. JSLS, 2014; 18: pii: e2014.00335.

Maruschke

, Kram

, Zimpfer

, Kundt

, Hakenberg

. Upper urinary tract tumors: Which diagnostic methods are needed?. Urol Int, 2017; 98:304–311.

Grahn

, Melle-Hannah

, Malm

, et al. Diagnostic accuracy of computed tomography urography and visual assessment during ureterorenoscopy in upper tract urothelial carcinoma. BJU Int, 2017; 119:289–297.

Golan

, Nadu

, Lifshitz

. The role of diagnostic ureteroscopy in the era of computed tomography urography. BMC Urol, 2015; 15:74.

Sadow

, Wheeler

, Kim

, Ohno-Machado

, Silverman

. Positive predictive value of CT urography in the evaluation of upper tract urothelial cancer. AJR Am J Roentgenol, 2010; 195:W337–W343.

10.

Matsumoto

, Naya

, Tobe

, et al. The usefulness of ureterorenoscopic examination in evaluation of upper tract disease. Int J Urol, 2006; 13:509–514.

11.

Vishal Vashistha

, Shabsigh

, Zynger

. Utility and diagnostic accuracy of ureteroscopic biopsy in upper tract urothelial carcinoma. Arch Pathol Lab Med, 2013; 137:400–407.

12.

Lim

, Shattuck

, Cook

. Pyelovenous lymphatic migration of transitional cell carcinoma following flexible ureterorenoscopy. J Urol, 1993; 149:109–111.

13.

Grasso

, McCue

, Bagley

. Multiple urothelial recurrences of renal cell carcinoma after initial diagnostic ureteroscopy. J Urol, 1992; 147:1358–1360.

14.

Liberati

, Altman

, Tetzlaff

, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: Explanation and elaboration. PLoS Med, 2009; 6:e1000100.

15.

Centre for Evidence-Based Medicine.. Oxford Centre for Evidence-based Medicine – Levels of Evidence (March 2009). www.cebm.net/2009/06/oxford-centre-evidence-based-medicine-levels-evidence-march-2009/ (accessed May 31, 2018).

16.

Parmar

, Torri

, Stewart

. Extracting summary statistics to perform meta-analyses of the published literature for survival endpoints. Stat Med, 1998; 17:2815–2834.

17.

Hozo

, Djulbegovic

, Hozo

. Estimating the mean and variance from the median, range, and the size of a sample. BMC Med Res Methodol, 2005; 5:13.

18.

Lee

, Kim

, Park

, et al. Correlation between the timing of diagnostic ureteroscopy and intravesical recurrence in upper tract urothelial cancer. Clin Genitourin Cancer, 2016; 14:e37–e41.

19.

Liu

, Su

, Xiong

, Li

, Zhou

. Diagnostic ureteroscopy for upper tract urothelial carcinoma is independently associated with intravesical recurrence after radical nephroureterectomy. Int Braz J Urol, 2016; 42:1129–1135.

20.

Nison

, Roupret

, Bozzini

, et al. The oncologic impact of a delay between diagnosis and radical nephroureterectomy due to diagnostic ureteroscopy in upper urinary tract urothelial carcinomas: Results from a large collaborative database. World J Urol, 2013; 31:69–76.

21.

Yoo

, You

, Song

, et al. Risk of Intravesical recurrence after ureteroscopic biopsy for upper tract urothelial carcinoma: Does the location matter. J Endourol, 2017; 31:259–265.

22.

Sankin

, Tin

, Mano

, et al. Impact of ureteroscopy before nephroureterectomy for upper tract urothelial carcinoma on oncologic outcomes. Urology, 2016; 94:148–153.

23.

Sung

, Jeon

, Han

, et al. Diagnostic ureterorenoscopy is associated with increased intravesical recurrence following radical nephroureterectomy in upper tract urothelial carcinoma. PLoS One, 2015; 10:e0139976.

24.

Ishikawa

, Abe

, Shinohara

, et al. Impact of diagnostic ureteroscopy on intravesical recurrence and survival in patients with urothelial carcinoma of the upper urinary tract. J Urol, 2010; 184:883–887.

25.

, Hendin

, Streem

S. H

, Klein

, Novick

. Impact of diagnostic ureteroscopy on long-term survival in patients with upper tract transitional cell carcinoma. J Urol, 1999; 163:783–785.

26.

Luo

, Kang

, Chen

, et al. Diagnostic ureteroscopy independently correlates with intravesical recurrence after nephroureterectomy for upper urinary tract urothelial carcinoma. Ann Surg Oncol, 2013; 20:3121–3126.

27.

Kulp

, Bagley

. Does flexible ureteropyeloscopy promote local recurrence of transitional cell carcinoma?. J Endourol, 1994; 8:111–113.

28.

Minowada

, Homma

, Takeuchi

, Kamijo

, Kitamura

. Long-term outcome of endoscopic biopsy and subsequent nephroureterectomy for upper urinary tract tumor. Int J Urol, 2001; 8:6–9.

29.

Seisen

, Granger

, Colin

, et al. A Systematic Review and Meta-analysis of Clinicopathologic factors linked to intravesical recurrence after radical nephroureterectomy to treat upper tract urothelial carcinoma. Eur Urol, 2015; 67:1122–1133.

30.

Yuan

, Chen

, Liu

, et al. Risk factors for intravesical recurrence after radical nephroureterectomy for upper tract urothelial carcinoma: A meta-analysis. Urol Oncol, 2014; 32:989–1002.

31.

Xylinas

, Colin

, Audenet

, et al. Intravesical recurrence after radical nephroureterectomy for upper tract urothelial carcinomas: Predictors and impact on subsequent oncological outcomes from a national multicenter study. World J Urol, 2013; 31:61–68.

32.

Pignot

, Colin

, Zerbib

, et al. Influence of previous or synchronous bladder cancer on oncologic outcomes after radical nephroureterectomy for upper urinary tract urothelial carcinoma. Urol Oncol, 2014; 32:23 e21–e28.

33.

Nissar

, McLuckie

, Anbarasan

, Biyani

, Nabi

. Ureteroscopic biopsy for suspected upper tract urothelial carcinoma is associated with increased intravesical recurrences at short follow-up: A multi institutional SUTURE group study. J Endourol, 2017; 30:322–323.Conference: 34th world congress of endourology, WCE 2016. South africa. Conference start: 20161108. Conference end: 20161112. http://onlinelibrary.wiley.com/o/cochrane/clcentral/articles/491/CN-01293491/frame.html.

34.

O'Brien

, Ray

, Singh

, Coker

, Beard

. British Association of Urological Surgeons Section of O: Prevention of bladder tumours after nephroureterectomy for primary upper urinary tract urothelial carcinoma: A prospective, multicentre, randomised clinical trial of a single postoperative intravesical dose of mitomycin C (the ODMIT-C Trial). Eur Urol, 2011; 60:703–710.

35.

Yamashita

, Ito

, Mitsuzuka

, et al. Clinical implications of intravesical recurrence after radical nephroureterectomy for upper urinary tract urothelial carcinoma. Int J Urol, 2016; 23:378–384.

36.

Keeley

, Kulp

, Bibbo

, McCue

, Bagley

. Diagnostic accuracy of ureteroscopy biopsy in upper tract transtional cell carcinoma. J Urol, 1997; 156:33–37.

37.

Roscigno

, Brausi

, Heidenreich

, et al. Lymphadenectomy at the time of nephroureterectomy for upper tract urothelial cancer. Eur Urol, 2011; 60:776–783.

38.

Furuse

, Matsushita

, Yajima

, et al. Systematic regional lymph node dissection for upper tract urothelial carcinoma improves patient survival. Jpn J Clin Oncol, 2017; 47:239–246.

39.

Seisen

, Colin

, Rouprêt

. Risk-adapted strategy for the kidney-sparing management of upper tract tumours. Nat Rev Urol, 2015; 12:155–166.