Routine Surveillance of Upper Urinary Tract Imaging for Diagnosing Upper Urinary Tract Urothelial Cancer Recurrence in Patients with Nonmuscle Invasive Bladder Cancer

Background Although routine surveillance imaging to examine upper urinary tract urothelial cancer recurrence during follow-up of nonmuscle invasive bladder cancer is recommended, its necessity remains invalidated. A single-institute long-term follow-up cohort study to evaluate the clinical impact of routine surveillance imaging and identify risk factors for upper urinary tract urothelial cancer recurrence after nonmuscle invasive bladder cancer treatment was conducted. Methods and Materials A retrospective chart review of 864 patients with primary nonmuscle invasive bladder cancer who underwent initial transurethral resection of bladder tumor between 1980 and 2020 was conducted. The opportunities to diagnose its recurrence were examined. Moreover, oncological outcomes included upper urinary tract urothelial cancer recurrence-free survival and overall survival. Results Of 864 patients, 19 (2.2%) experienced upper urinary tract urothelial cancer recurrence. Among 19 patients, recurrence was detected through routine imaging in 12 (63.2%), cystoscopy in 2 (10.5%), urine cytology in 2 (10.5%), and presence of gross hematuria in 1 (5.3%). All patients had high- or highest-risk NMIBC at diagnosis of primary nonmuscle invasive bladder cancer. On multivariate Fine-Gray proportional regression analyses, a tumor size of ≥30 mm and carcinoma in situ were independently associated with short upper urinary tract urothelial cancer recurrence-free survival (P=0.040 and 0.0089, respectively). Conclusion Most patients experiencing upper urinary tract urothelial cancer recurrence were diagnosed by routine surveillance imaging, suggesting its clinical importance, especially for patients with nonmuscle invasive bladder cancer accompanied by a tumor size of ≥30 mm and carcinoma in situ.

Each guideline mentions the need for follow-up for NMIBC with regards to the observation of the upper urinary tract.Table 1 summarizes the follow-up strategies for the diagnosis of UTUC using upper urinary tract imaging.Routine surveillance imaging, including computed tomography urography (CTU), is recommended for patients with high-risk NMIBC but not for those with low or intermediate risk [11][12][13][14][15].Although CTU has excellent diagnostic performance for UTUC, a risk of contrast nephropathy and radiation exposure exists.Sternberg et al. reported that most UTUC recurrences in patients with NMIBC were missed on routine surveillance imaging, which suggests that this method may not be efcient to diagnose UTUC recurrence [16].Some authors reported that routine investigations detected 38-78% of UTUC recurrences, even in cohorts of patients after cystectomy [17][18][19].Te necessity of routine surveillance imaging for diagnosing UTUC after treating bladder cancer remains unclear.If routine imaging for UTUC recurrence is not needed, as described by Steinberg et al., expensive and radiation-exposed examinations that may have adverse efects on both the healthcare system and the human body should be avoided.A single-institute, long-term follow-up cohort study to determine the necessity of routine surveillance imaging in a real-world clinical setting and to identify the risk factors for UTUC recurrence after NMIBC treatment was conducted.

Patient Selection and Study Design.
Te study protocol was approved by the Institutional Review Board for Clinical Studies (Medical Ethics Committee ID: 2891) at the Nara Medical University in Nara, Japan.Te opt-out method was applied to obtain consent from the participants via posters and/or websites, and the study was conducted in compliance with the principles of the Declaration of Helsinki.A preprint of this manuscript has previously been published [20].
A retrospective chart review of 913 patients diagnosed with primary NMIBC who underwent transurethral resection of bladder tumor (TURBT) between 1980 and 2020 was conducted.Te inclusion criteria were as follows: (1) primary NMIBC at diagnosis at the Nara Medical University, (2) UTUC recurrence after initial TURBT, and (3) complete clinicopathological data.
Te following clinicopathological variables on prognosis were evaluated: age, sex, Eastern Cooperative Oncology Group performance status (ECOG-PS), tumor size, multiplicity, clinical T category, World Health Organization (WHO) 1973 and 2004 grades, CIS, prostate-involving CIS, lymphovascular invasion (LVI), variant histology, risk classifcation, and intravesical recurrence after initial TURBT.Te Clinical T category was determined on imaging before TURBT and through the pathological diagnosis of TURBT.After the advent of the WHO 2004 schemes, the pathologists at our hospital used the WHO 2004 grade instead of the WHO 1973 grade.Te risk classifcation of NMIBC was defned in the Clinical Practice Guidelines for Bladder Cancer 2019 [14].Te criteria for highest-risk NMIBC are as follows: (1) T1 and high grade (HG) tumors with concomitant bladder CIS, concomitant prostate-involving CIS, multiple tumors, recurrent tumors, a tumor size of ≥30 mm, variant histology, and/or LVI and (2) BCG-unresponsive disease.BCG-unresponsive disease was also defned in the Clinical Practice Guidelines for Bladder Cancer 2019 [14].
2.2.Follow-Up.Patients were followed up after the initial TURBT according to the protocols of the Nara Medical University.Cystoscopy was performed three months after the initial TURBT.Routine cystoscopy, blood examination, and urine cytology for diagnosing intravesical recurrence were performed every three, six, or 12 months for up to ten years according to the clinicopathological factors.To diagnose the UTUC recurrence, routine radiographic imaging examinations were performed every 6 or 12 months after initial TURBT and further intensifed by each attending physician according to the patient's symptoms.

Outcomes.
All opportunities to diagnose UTUC recurrence were examined.Oncological outcomes included the UTUC recurrence-free survival and overall survival (OS).UTUC recurrence-free survival OS were defned as the duration from initial TURBT.

Statistical Analysis.
Quantitative variables are summarized as means (standard deviation (SD)), and categorical variables are presented as proportions.Student's t-test, Fisher's exact test, and the chi-square test were used as appropriate for the statistical analyses of patient clinicopathological characteristics.Prognostic factors for UTUC recurrence-free survival were assessed using a competing risk analysis by the Gray test for cumulative incidence.Competing events were defned as deaths without UTUC recurrence.Univariate and multivariate Fine-Gray competing risk regression analyses were used to estimate the hazard ratios for UTUC recurrence-free survival.Te OS was assessed using Kaplan-Meier survival curves constructed using GraphPad Prism 7.0 (GraphPad Software, San Diego, CA, USA).Survival curves were compared using the log-rank test.A p value <0.05 was considered statistically signifcant.Statistical analyses were performed using EZR (Saitama Medical Center, Jichi Medical University, Saitama, Japan) [21].

Characteristics of Patients with UTUC Recurrence.
Figure 1 presents the fowchart of the creation of the patient cohort dataset.Forty-four patients were excluded due to lack of data, and fve patients were excluded as they were diagnosed with UTUC at the same time as primary NMIBC.Ultimately, 864 patients were included in the analysis.Table 2 shows the clinicopathological characteristics of patients and a comparison between the UTUC recurrence and nonUTUC recurrence groups.Of the 864 patients, 19  ), one had pTis (5.9%), and two had pTa (11.8%).One patient received neoadjuvant chemotherapy and was pathologically diagnosed with ypT0.Tree patients were diagnosed with a UTUC recurrence more than ten years after initial TURBT.

Prognostic
Factors for UTUC Recurrence. Figure 3 and Supplementary Figure S1 show the cumulative incidence of UTUC recurrence-free survival after initial TURBT according to the clinicopathological variables.Figure 3 demonstrates that a tumor size of ≥30 mm, high grade, CIS-positive NMIBC, cTis or cT1, and high-or highestrisk NMIBC were associated with a shorter UTUC recurrence-free survival.Figure 3(d) shows that there was a signifcant diference in UTUC recurrence for each T category.Furthermore, Figure 3(e) shows no signifcant diference in UTUC recurrence between high-and highest-risk NMIBC groups.However, high-or highestrisk NMIBC group was signifcantly associated with a shorter UTUC recurrence-free survival than the low-and intermediate-risk NMIBC group.Table 4 shows the univariate and multivariate Fine-Gray competing risk regression analyses for UTUC recurrence-free survival.Figure 4 shows the cumulative incidence of UTUC recurrence-free survival in the four groups categorized according to the tumor size and CIS status.Patients with both a tumor size of ≥30 mm and CIS-positive NMIBC had a signifcantly shorter UTUC-recurrence-free survival than those lacking these two variables.Multivariate analyses revealed that a tumor size of ≥30 mm (hazard ratio (HR): 2.53; 95% confdence interval (CI): 1.04-6.15;and Fine-Gray P � 0.040) and CIS-positive NMIBC (HR: 4.78; 95% CI: 1.48-15.42;and Fine-Gray P � 0.0089) were independently associated with a short UTUC recurrencefree survival.[11] Unnecessary Every 1-2 years Every 1-2 years CUA [12] Not mentioned Not mentioned Every 1-2 years EAU [13] Not mentioned Not mentioned Every 1 years JUA [14] As clinically indicated As clinically indicated Every year for up to 3 years and thereafter every 2 years NCCN [15] As       We have summarized our original devised follow-up protocol for UTUC recurrence.For high-risk NMIBC, the authors recommend that imaging examinations, including CTU, are conducted annually until the ffth year.For high-risk NMIBC with a tumor size of ≥30 mm and CIS, the authors recommend six-month intervals of monitoring up to the ffth year.Routine surveillance is unnecessary for low-and intermediate-risk NMIBC, but noninvasive examinations such as ultrasonography and urine cytology may be performed as needed.

Advances in Urology
Advances in Urology

Discussion
It has been reported that the routine surveillance imaging for diagnosing UTUC recurrence after the treatment with NMIBC is not necessary [16,22].Sternberg et al. reported that routine imaging identifed UTUC recurrences in 4 (25%) out of 16 patients with Ta bladder cancer and 9 (27%) out of 33 patients with T1 bladder cancer.Tere were no diferences in the patient cohort regarding deviation of T category compared to our cohort.In contrast to their fndings, our study showed that among 19 patients with UTUC recurrence, the majority (84.2%) were diagnosed by routine surveillance imaging, cystoscopy, or urine cytology and only one (5.3%)was diagnosed by the presence of symptoms (Figure 2 and Table 3).Tis discrepancy in these results might potentially be attributed to diferences of physician's awareness, with a tendency for patients with T1 tumors to be received more cautiously than those with Ta tumors.Moreover, several cases were found to present with invasive urothelial cancer, and 9 cases (47.4%) in the UTUC recurrence group had progressed to muscle invasion disease.If routine surveillance imaging was not performed in patients, unresectable or metastatic disease may have developed.Moreover, UTUC recurrence increased all-cause mortality (Figure 5(b)) and, therefore, routine surveillance imaging might be necessary to reliably detect upper urinary tract recurrence at an early stage.Most guidelines recommend imaging examination to detect UTUC recurrence for high-risk NMIBC, while they suggest no such procedure for low-or intermediate-risk NMIBC.However, there are few reports to actually evaluate the follow-up strategy for detecting UTUC recurrence.In such a situation, our analysis shows the same conclusion because all patients in the UTUC recurrence group had high-or highest-risk NMIBC, supporting the guidelines' recommendation.However, it is also considered that it may not be applicable to all patients.Te overall incidence of UTUC recurrence in our study was very low at 2.2%, which was almost the same as that of previous reports [1][2][3][4], and all patients with UTUC recurrence had high-or highest-risk NMIBC (Table 3).In other words, among patients with highor highest-risk NMIBC, only a handful does not experience UTUC recurrence.Among these patients, criteria are needed to select those requiring more careful follow-up.Figures 4  and 5 show that a tumor size of ≥30 mm and CIS-positive NMIBC increased the UTUC recurrence and all-cause mortality.Previous reports also described that a history of CIS-positive NMIBC developed aggressive UTUC and increased the cancer-related mortality rate after radical nephroureterectomy for UTUC [23].Consequently, it is advisable to use tumor size and CIS as these criteria.Even when all patients with high-risk NMIBC undergo imaging examinations of the upper urinary tract, the detection rate for UTUC recurrence remains low.Terefore, we believe it is crucial to pay close attention to those with the highest-risk NMIBC.
In contrast, there were no patients with low-and intermediate-risk NMIBC who experienced a UTUC recurrence (Table 3).Previous reports have also described the similar results [7][8][9][10].For patients with low-or intermediate-risk NMIBC, follow-up with ultrasonography and urine cytology, which are non-invasive, without radiographic imaging, such as CTU, might be sufcient and cost-saving.Our thoughts on the follow-up protocol for UTUC recurrence are summarized in Figure 6.For high-risk NMIBC, imaging examinations, including CTU, should be conducted annually until the ffth year, as is customary.As previously mentioned, a tumor size of ≥30 mm and CIS were found to be signifcantly independent risk factors for UTUC recurrence, and thus we believe that the management of the highest-risk NMIBC cases with a tumor size of ≥30 mm and CIS warrants careful consideration.Furthermore, for highrisk NMIBC patients with a tumor size of ≥30 mm and CIS, we recommend six-month intervals of monitoring up to the ffth year rather than twelve-month intervals recommended by many guidelines.On the other hand, the accuracy of CTU and urine cytology are limited [11][12][13].DNA methylation analysis can be recently utilized to improve the accuracy to detect UTUC recurrence [24,25].In the case of UTUC detection, samples of blood and urine are used, from which DNA is extracted and the methylation patterns of specifc gene regions within the DNA are analyzed.Tis approach may ofer superior accuracy to CTU or urine cytology, with less invasiveness for patients than ureteroscopy.Incorporating this innovative analysis in the future could be signifcantly simplify the detection of UTUC recurrence.
Tis study has several limitations.First, data on patients were collected retrospectively, and a potential selection bias may have occurred.For example, this cohort included cases from 1980 onwards.To date, the diagnostic abilities of radiographic imaging and endoscopic techniques have advanced, and the follow-up has not been constant between cases.Tere is a signifcant diference in image quality and resolution of CT scans with improvements in image reconstruction techniques.Lower image quality of CT scans may result in missing of small lesions.In addition, the decrease of radiation exposure and the cost of examinations have alleviated the burden on patients undergoing tests.Tis might afect the ability to detect UTUC recurrences because the follow-up protocol intervals have varied between 1980 and 2020.It is challenging to accurately evaluate these data unless these intervals are not standardized.Second, the cohort of UTUC recurrence was too small and there was the signifcant disparity in patient numbers between the UTUC recurrence and nonUTUC recurrence groups.Among 19 patients in the UTUC recurrence group, it is possible that a normal distribution was not observed in either variable due to the limited sample size.Furthermore, the impact of individual cases becomes magnifed in such a small cohort.For instance, if only one out of the 19 patients was detected by the diferent examination, the recurrence rate could change to 5.3%, which is a signifcant alteration that could impact the comprehension of these data.Tird, due to insufcient data, we were unable to categorize patients into more appropriate groups as follows: those receiving routine surveillance imaging and those not receiving it.As a result, all patients did not undergo this imaging, potentially introducing a signifcant selection bias.Fourth, intravesical 10 Advances in Urology BCG therapy mostly depends on each attending physician and is not uniformly administered for CIS-positive and high-risk NMIBC.Moreover, the number of cases with UTUC recurrence was so small that the HR of intravesical BCG therapy after TURBT could not be calculated, and its impact could not be assessed.Fifth, since multiple surgeons performed TURBT, prognosis might have been infuenced by diferences in skill.Further prospective cohort studies are required to substantiate the fndings of this study.Sixth, there is the uncertain association between vesicoureteral refex (VUR) and UTUC recurrence.Resecting tumors located around ureteral orifce during TURBT may inadvertently disrupt the vesicoureteral junction, potentially causing the migration of cancer cells to the upper urinary tract.However, due to in sufcient data regarding the resection area during TURBT, we are unable to evaluate this relationship.

Figure 5
compares the survival curves for OS among patients stratifed according to the number of risk factors for UTUC recurrence (A) and between those with or without UTUC recurrence (B).Prognostic factors included a tumor size of ≥30 mm and CIS-positive NMIBC.Patients with both prognostic factors had signifcantly shorter OS (log-rank P � 0.042) than those without.Moreover, UTUC recurrence was signifcantly associated with a shorter OS (HR: 3.95; 95% CI: 1.15-13.65;and log-rank P � 0.030).

Figure 3 :
Figure 3: Cumulative incidence of UTUC recurrence-free survival after initial TURBT according to the tumor size, WHO 2004 grade, CIS, clinical T category, and risk classifcation.Survival curves of the cumulative incidence of UTUC recurrence-free survival after initial TURBT for primary NMIBC were plotted according to the tumor size (a), WHO 2004 grade (b), CIS (c), clinical Tcategory (d), and risk classifcation (e) using the Gray test.† Te risk was classiNed according to the Clinical Practice Guidelines for Bladder Cancer 2019 [14].

11 P = 0. 14 P <0. 001 Figure 4 :Figure 5 :
Figure4: Te cumulative incidence of UTUC recurrence-free survival after initial TURBT according to tumor size and CIS.Survival curves of the cumulative incidence of UTUC recurrence-free survival after initial TURBT for primary NMIBC were plotted according to the tumor size and CIS.Patients are categorized into four groups based on whether the tumor size is 30 mm or larger and the presence of CIS.

Figure 6 :
Figure6: An original follow-up protocol for UTUC recurrence.We have summarized our original devised follow-up protocol for UTUC recurrence.For high-risk NMIBC, the authors recommend that imaging examinations, including CTU, are conducted annually until the ffth year.For high-risk NMIBC with a tumor size of ≥30 mm and CIS, the authors recommend six-month intervals of monitoring up to the ffth year.Routine surveillance is unnecessary for low-and intermediate-risk NMIBC, but noninvasive examinations such as ultrasonography and urine cytology may be performed as needed.

Table 1 :
Follow-up strategies for diagnosing UTUC after treatment for NMIBC by upper urinary tract imaging.

Table 2 :
Clinicopathological characteristics and comparison between the UTUC recurrence group and the non-UTUC recurrence group.
Figure1: Flowchart of the creation of the patient's cohort dataset.A retrospective chart review of 913 patients diagnosed with primary NMIBC who underwent TURBT between 1980 and 2020 was conducted.Forty-four patients were excluded due to lack of data, and fve patients were excluded because they had UTUC at the time of diagnosis with primary NMIBC.Ultimately, 864 patients were included in the analysis.
[14] Bacillus Calmette-Guérin; CIS: carcinoma in situ; ECOG-PS: Eastern Cooperative Oncology Group performance status; IPIC: immediate postoperative instillation of chemotherapy; LVI: lymphovascular invasion; SD: standard deviation; TURBT: transurethral resection of bladder tumor; UTUC: upper urinary tract urothelial carcinoma; WHO: World Health Organization.†Terisk was classifed according to the Clinical Practice Guidelines for Bladder Cancer 2019[14].Figure 2: Pie chart of the opportunities to diagnose UTUC recurrence.Te opportunities to diagnose UTUC recurrence in UTUC recurrence group are drawn.Tese are divided into routine imaging, cystoscopy, urine cytology, gross hematuria and unknown.

Table 3 :
A summary of the opportunities to diagnose UTUC recurrence, clinicopathological characteristics of primary NMIBC, and pathological T categories of UTUC.

Table 4 :
Univariate and multivariate Fine-Gray competing risk regression analysis for UTUC recurrence-free survival in patients with primary NMIBC.