Upper tract urothelial carcinoma (UTUC) is a urothelial cancer that occurs in the ureter or pelvocaliceal system. It has an incidence of 1–2 cases per 100,000 patients (1, 2). Recently, the incidence of UTUC has been increasing with the development of diagnostic techniques, such as radiological and endoscopic techniques (3). Although the etiology of UTUC remains unclear, cigarette smoking, herbal medicines, chronic infection, and occupational carcinogenesis are known risk factors for UTUC (3). Its gold standard treatment is radical nephroureterectomy with bladder cuff excision (RNUx), which has excellent oncological outcomes (4). However, an important disadvantage of RNUx is intravesical recurrence (IVR), which occurs in 15%–50% of patients after RNUx (5, 6). Diagnosis of IVR after RNUx is important because additional treatments, such as surgery and chemotherapy, due to bladder cancer are required. Moreover, it may affect the prognosis of UTUC (7).

UTUC and bladder cancer have the same histological subtype and urothelial composition. Nevertheless, the two have embryological, epidemiological, and molecular differences (8). Because of these common features, approximately 13% of patients with UTUC have a history of bladder cancer, and approximately 9% of UTUC cases are diagnosed at the same time as bladder cancer (9). Furthermore, there is a high IVR rate after RNUx (10).

Accurate assessment of risk factors for IVR after RNUx is important in reducing unnecessary examinations and treatments for bladder cancer (11). Several previous studies were conducted with the aim of evaluating and preventing IVR, which occurred after RNUx. In these studies, tumor location, cancer stage, grade, and sex were reported as risk factors for IVR after RNUx (7, 12, 13). However, there are limited guidelines on the follow-up protocol for IVR. Despite the limited evidence, IVR occurs most commonly within 12 months after RNUx (14). Moreover, there are insufficient reports on the probability of recurrence of IVR or follow-up when there is no recurrence within 12 months or no IVR for a specific period. In particular, there is no report of IVR-free (IVRF) survival for patients who did not have a history of bladder cancer or concomitant bladder cancer and UTUC.

Therefore, this study aimed to evaluate the conditional IVRF survival rate of patients with UTUC who had no history of bladder cancer and no concomitant bladder cancer and to determine the incidence and risk factors for IVR by analyzing a relatively large number of patients who underwent RNUx for UTUC.

Among the 1,095 patients, 462 patients developed IVR. The mean age of patients with and without IVR was 64.88 ± 10.35 and 65.92 ± 11.28 years, respectively (p = 0.120). The prevalence rates of HTN and DM were 48.05% and 22.29%, respectively, in the IVR group and 41.71% and 16.43%, respectively, in the without IVR group (HTN: p = 0.037, DM: p = 0.014). A total of 30.74% patients in the IVR group and 15.32% in the without IVR group had a history of bladder cancer (p < 0.001). Multifocal tumors were observed in 25.54% patients in the IVR group and 22.75% in the without IVR group (p = 0.003).

Cuffing was performed using the intravesical approach in 51.95% patients in the IVR group and 60.19% in the without IVR group (p = 0.007). Preoperative URS was performed in 59.53% patients in the IVR group and 44.39% in the without IVR group (p < 0.001). Adjuvant chemotherapy was performed in 81 patients (17.53%) in the IVR group and 159 patients (25.12%) in the without IVR group (p < 0.001) (Table 1).

IVR occurred at a mean period of 13.08 (interquartile range, 3.97–14.05) months after RNUx. The median IVRF survival period was 12.30 (95% CI, 7.72–16.88) months in patients with a history of bladder cancer or concomitant bladder cancer and 133.00 (95% CI, 57.89–208.11) months in patients without such a history (p < 0.001) (Figure 1). Multivariable analysis revealed a history of bladder cancer [hazard ratio (HR) 2.409, p < 0.001], a multifocal tumor (HR 1.348, p = 0.008), use of preoperative URS (HR 1.733, p < 0.001), extravesical bladder cuffing (HR 1.408, p = 0.009), LN involvement (HR 2.121, p = 0.004), positive surgical margins (HR 1.553, p = 0.026), and adjuvant chemotherapy (HR 0.759, p = 0.033) to be risk factors for IVR (Table 2).

In this study, the risk factors for IVR after RNUx in patients with UTUC are discussed. Moreover, it was confirmed that there were no significant UTUC-related risk factors for IVR among patients who had ≥36-month IVRF survival rate. However, it was confirmed that approximately 5% of patients with an IVRF period of ≥36 months developed IVR.

After RNUx for UTUC, the cause of IVR is not clear yet. However, the possible mechanisms of IVR include the field theory that UTUC is exposed to the urothelium to generate IVR (16) and the intraluminal tumor seeding theory that the cancer cells in the upper tract reach the bladder via the urinary stream (17). If URS is performed before RNUx, the intraluminal pressure of the renal pelvis increases and tumor manipulation occurs, resulting in increased intraluminal tumor seeding and eventually increased IVR (18, 19). In the present study, it was found that URS before RNUx was a risk factor for IVR in patients with 0–24-month IVRF survival rate. The efficacy of URS before surgery remains controversial (20, 21). However, reducing the use of preoperative URS based on radiological imaging and laboratory examinations, such as urine analysis and cytology, could improve the IVR rate.

In a recent meta-analysis, the predictors of IVR were classified into three categories: patient-specific factors, tumor-specific factors, and treatment-specific factors (22). In this study, patient-specific factors included male sex, bladder cancer history, and chronic kidney disease; tumor-specific factors included positive urinary cytology, ureteral location, multifocality, invasive pT stage, and necrosis. Furthermore, treatment-specific factors included laparoscopic surgery, extravesical bladder cuff excision, and positive surgical margins. In the present study, bladder cancer history, preoperative URS, multifocality, LN invasion, bladder cuffing, and surgical margin involvement were identified as risk factors for IVR after RNUx. The present study was also conducted in patients with no history of bladder cancer. In these patients, DM, gross hematuria, URS, and extravesical bladder cuff excision were evaluated as predictive factors of IVR after RNUx possibly owing to the difference between the groups. However, it is clear that avoiding preoperative URS and extravesical bladder cuff excision can reduce IVR.

Previous studies reported that DM is an independent risk factor for IVR (23, 24). Moreover, a previous study reported that DM corresponded to a poor prognosis in patients with UTUC. Chronic exposure to hyperinsulinemia or hyperglycemia is a possible factor that induced tumor cell proliferation (25, 26). Hashimoto et al. (27) reported that gross hematuria was a significant risk factor for IVR. They suggested that hemorrhagic tumor cells could be seeded easily in the mucosal epithelium. Nevertheless, the methods of management for bladder cuff remain controversial. Some studies reported that the method of bladder cuff excision was not associated with the IVR rate (28, 29). However, when intramural ureter was not appropriately resected, the reported recurrence rate was 33%–75% (30).

Recently, Katims et al. (31) reported the risk factors for IVR after minimally invasive RNUx. They reported that IVR occurred in 22.7% of patients who underwent RNUx. In addition, URS, transurethral resection of the bladder cuff and positive surgical margin were suggested to be risk factors for IVR. In the present study, 37.38% of patients without prior or concurrent bladder cancer developed IVR. The different cancer stage of the enrolled patients might have affected higher IVR rate.

Due to concerns about IVR, bladder examination, including cystoscope and urinary cytology, is recommended for 5 years after RNUx for UTUC (4). However, due to the lack of studies, optimal follow-up strategies for IVR after RNUx cannot be concluded (32, 33). Recently, Shigeta et al. (34) reported on conditional IVRF survival after RNUx of the 364 patients with Ta-T3 UTUC. According to this study, IVR was identified in 48.4% patients; the 5-year conditional IVRF survival rate increased from 41.5% to 60.5%, 73.4%, 79.5%, and 96.7% in patients with 1-, 2-, 3-, and 4-year IVRF survival rate, respectively. In the present study, each conditional IVRF survival was evaluated for patients with 0.5-, 1-, 2-, 3-, 4-, and 5-year IVRF survival rate. Our results are consistent with those of a previous study. Moreover, gross hematuria, URS, retroperitoneal approach, and a low grade were identified as risk factors for IVR in patients with 1-year IVRF survival rate, and only URS was evaluated as a risk factor for IVR in patients with 2-year IVRF survival rate. In addition, for patients with >3-year IVRF survival rate, these risk factors were deemed to have no significant effect on IVR. However, IVR occurred in 13.1% of patients with 3-year IVRF survival rate, and only 5.6% of patients with 5-year IVRF survival rate were diagnosed with IVR. Although these could not be clearly identified as IVR associated with UTUC, it was found to be a higher incidence than that in the general population. Therefore, patients with UTUC require IVR follow-up even following 5 years after RNUx.

Conditional survival analysis is a method of assessing additional survival at a specific time point after initial diagnosis or treatment (35). It is widely performed in cancer research because it can transmit additional important and diverse information during follow-up (36, 37). In this study, conditional IVRF survival according to the IVRF period was found in a large number of UTUC patients without a history of bladder cancer. According to the conditional IVRF survival rate in the present study, in the case of patients whose IVRF period is <36 months, the IVR rate after the IVRF period is >10%; hence, active assessment is required until 3 years after RNUx. In addition, since IVR occurs in 3%–5% of patients with an IVRF period of ≥36 months, patient education and screening tests, such as regular urine analysis and cytology, are required.

This study has several limitations. First, it was a retrospective study; thus, fixed criteria for diagnosis, treatment, and patient follow-up could not be used. Second, the indications for the use of URS, lymphadenectomy, and adjuvant chemotherapy were not clear. Although this study could not suggest an optimal follow-up strategy for IVR after RNUx for UTUC, it was considered that follow-up for IVR could be optimized through systematic analysis if additional data would be collected in the future. Moreover, notably, to the best of our knowledge, it is the first large-scale study on the conditional survival of patients with UTUC who had no history of bladder cancer.

In this study, a history of bladder cancer, multifocal tumors, preoperative URS, LN invasion, extravesical bladder cuffing, and surgical margin involvement were identified as risk factors for IVR. However, in patients with no history of bladder cancer, DM, gross hematuria, preoperative URS, and extravesical bladder cuffing were risk factors. In patients whose IVRF period is <3 years, the IVR rate is ≥10%; hence, active IVR assessment is required until 3 years after RNUx. In addition, since 3%–5% of patients with an IVRF survival period of ≥3 years develop IVR, patient education and regular screening tests, such as urine analysis and cytology, are required for patients with ≥3-year IVRF survival rate.

The raw data supporting the conclusions of this article will be made available by the authors without undue reservation.

The studies involving human participants were reviewed and approved by Samsung medical center. Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements.

HS contributed to the conceptualization. WS, MK, HJ, BJ, SS, SJ, and HL contributed to the methodology. JC contributed to the formal analysis. JC and HS contributed to the data curation. JC contributed to writing–original draft preparation. HS contributed to writing–review and editing and supervision. All authors contributed to the article and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.