Long-Term Oncologic Outcomes of Endoscopic Management of High-Risk Upper Tract Urothelial Carcinoma: The Fundació Puigvert's Experience

Introduction

Owing to the refinement of endoscopic armamentarium, indications of renal sparing surgery have been progressively expanded over time and is now recommended ¹ as a primary treatment option for patients with low-risk upper tract urothelial carcinoma (UTUC) because of excellent long-term oncologic outcomes. ² Radical nephroureterectomy (RNU) with bladder cuff excision remains the gold standard for the treatment of nonmetastatic high-risk UTUC. ¹ However, many patients outside of the low-risk criteria may possess low absolute risks of distant progression.

Several large-scale retrospective cohorts based on RNU specimens have attempted to refine patient selection for an endoscopic approach, but they are limited by the use of a surrogate endpoint (i.e., ≥pT2 and/or N+ tumors), which limits the generalizability of the results since redirection of patients to RNU favored selection of aggressive disease. ^3,4 To date, there is a lack of long-term data on endoscopic treatment in patients with high-risk UTUC. Herein, we hypothesized that careful selection of high-risk patients undergoing an endoscopic approach could result in acceptable oncologic outcomes, as compared with those achieved after radical treatment in the same risk-category population.

Methods

Patients with high-risk UTUC managed endoscopically between 2015 and 2021 were retrospectively identified from a prospectively maintained single academic institution database (IRB approval: FP2014/17). High-risk disease was defined according to the European Association of Urology (EAU) criteria ¹ by the presence of any of the following factors: multifocal disease, tumor size ≥20 mm, high-grade cytology, high-grade ureteroscopy (URS) biopsy, local invasion on CT, hydronephrosis, previous radical cystectomy for high-grade bladder cancer, and variant histology.

Elective and imperative indications for endoscopic treatment were considered. Imperative indications included solitary kidney, severe chronic kidney disease (estimated glomerular filtration rate ≤30 mL/[min·m²]), bilateral disease, contraindication to RNU, and genetic predisposition. Regarding elective indications, we offered endoscopic treatment of high-risk tumors if the following criteria were met: 1/complete surgical removal was achieved; 2/no invasive appearance of the tumor on CT scan; and 3/no histological variant. The final decision was routinely validated at a multidisciplinary meeting.

Our endoscopic technique has been described previously. ⁵ In brief, careful exploration of the upper tract is systematically undertaken using a semirigid and then a digital flexible URS (Flex XC; Karl Storz, Germany, URF-V; Olympus, Japan) with the aid of image enhancement modalities (Image1S; Karl Storz, Narrow Band Imaging; Olympus). A 10F to 12F ureteral access sheath is then inserted with its upper end under the target lesion or the ureter pelvis junction.

According to lesion size and morphology, different biopsy devices were used, including 3F ureteroscopic forceps, 6F back-loading forceps (BIGopsy; Cook Medical), and a 2.2F Nitinol basket (N-Circle; Cook Medical). For tumors that could not be safely treated by pure retrograde access, a combined percutaneous approach was performed. Tumor ablation was performed with thulium–YAG laser (standard setting: 10–15 W). For large lesions, we used the combination of thulium and holmium–YAG lasers (standard setting: 0.8–1.2 J and 8–12 Hz), which allows removing any necrotic layer developed by the photothermic coagulative effect of thulium laser and so to show up any residual tumor tissue. ⁶

An example of the surgical technique is provided in Supplementary Video S1. At the end of the procedure, the ureteral access sheath was removed under endoscopic visual control to assess whether any ureteral injuries occurred, and a final retrograde pyelography was performed to detect any possible leakage. A 6F to 7F Double-J or Mono-J ureteral stent was placed and maintained for a variable timeframe according to the findings of the pyelography. Postoperative complications were graded according to Clavien–Dindo classification.

Surgical specimens were evaluated by a dedicated genitourinary pathologist (F.A.). All patients underwent strict follow-up including CT scan, flexible cystoscopy, and urine cytology every 6 months. Kaplan–Meier curves were used to illustrate oncologic outcomes after treatment: UTUC recurrence-free survival, bladder recurrence-free survival, RNU-free survival, metastasis-free survival (MFS), cancer-specific survival (CSS), and overall survival (OS). UTUC recurrence-free survival included only ipsilateral recurrence. The decision to perform RNU was usually based on grade progression, invasive aspect on CT scan, recurrence unfit for endoscopic treatment, or functional loss of the renal unit. Oncologic outcomes were compared between elective and imperative indications with the log-rank test. All statistical analyses were performed using R software version 4.1.3 (R Foundation for Statistical Computing, Vienna, Austria).

Results

A total of 60 patients with high-risk UTUC met our inclusion criteria. Baseline characteristics are summarized in Table 1. The indication for endoscopic management (Table 2) was considered imperative in 29 patients (48%) as a result of the presence of a single kidney in 11 patients (19%), severe chronic kidney disease in 6 patients (10%), bilateral disease in 5 patients (8%), contraindication to RNU in 5 patients (8%), and Lynch syndrome in 2 patients (3%). The remaining indications were elective (n = 31; 52%). Six (10%) patients experienced minor complications (Clavien–Dindo 2) and two (3.3%) major complications (Clavien–Dindo 3a and 3b) after treatment. The median follow-up in patients without any event was 36 months.

Figure 1 shows the Kaplan–Meier survival curves and Table 3 gives the estimated survival. Overall, three patients developed metastases over follow-up and 16 patients died, including 5 patients with disease-specific mortality. The estimated OS, CSS, and MFS were 94% (89–100), 98% (94–100), and 98% (94–100), respectively at 1 year, 75% (63–89), 95% (88–100), and 94% (87–100), respectively, at 3 years, and 57% (41–79), 75% (57–99), and 86% (71–100), respectively, at 5 years. A total of 18 patients experienced 26 upper tract recurrences over follow-up, most of them were treated endoscopically. The estimated UTUC recurrence-free survival was 83% (74–94) at 1 year, 70% (58–84) at 3 years, and 56% (40–76) at 5 years.

OPEN IN VIEWER

FIG. 1.

Kaplan–Meier curves. RNU = radical nephroureterectomy; UTUC = upper tract urothelial carcinoma. Color graphics are available online.

A total of nine patients proceeded to RNU over follow-up. The estimated RNU-free survival was 87% (79–96) at 1 year, 84% (75–95) at 3 years, and 81% (70–93) at 5 years. Finally, a total of 12 patients experienced bladder recurrence, all of whom were classified as nonmuscle-invasive bladder cancer. The estimated bladder recurrence-free survival was 87% (79–96) at 1 year, 80% (69–92) at 3 years, and 69% (54–88) at 5 years. All oncologic outcomes were similar between patients with elective and imperative indications (Table 4, all log-rank p-values >0.05).

Discussion

Endoscopic management of UTUC is an effective treatment in patients with low-risk disease. Conversely, patients with high-risk features and normal contralateral kidney are usually submitted to radical treatment since poorer oncologic outcomes are expected in high-grade, multifocal, or large tumors. Nevertheless, we hypothesized that not all high-risk patients are equal and that some of them could benefit from conservative endoscopic treatment. In this study, we provide sound evidence that endoscopic conservative management offers effective oncologic control with an acceptable CSS when high-risk patients with UTUC are properly selected.

In a recent systematic review and meta-analysis, Kawada et al. reported similar survival outcomes between RNU and endoscopic treatment in UTUC. ⁷ Although the majority of included patients had low-risk disease, up to 24% of endoscopically treated patients had multifocal tumors, and up to 68% had high-grade tumors. Similarly, despite all the patients included in our study being defined as having high-risk disease, only 30%, 38%, and 43% of patients had high-grade, multifocal, and large disease (>2 cm), respectively, and these patients achieved comparable oncologic outcomes with that reported in the literature after RNU.

According to these findings, it is clear that although the EAU risk classification helps in identifying those patients with more aggressive clinical behavior to be submitted to prompt radical treatment, a non-negligible portion of high-risk patients is possibly being overtreated.

Thus, our results corroborate that a subset of patients with high-risk UTUC may benefit from conservative treatment and that UTUC risk categories should be also refined. In this regard, the selection of high-risk patients for endoscopic management remains a matter of debate. The success of endoscopic management depends on the ability to identify preoperative features that distinguish patients with a low risk of progression who can be managed appropriately with endoscopic kidney-sparing surgery from those who will require definitive and radical treatment. ^3,8

Several large-scale retrospective studies have attempted to change the two-level currently recommended classification of UTUC. ^3,4 However, the exclusion of patients with UTUC who received a primary kidney-sparing surgery treatment partly limits the strength of their findings. Although not comparative, our study also provides some useful information regarding the selection criteria for endoscopic treatment in high-risk UTUC. Excluding patients with high-burden tumors, invasive CT appearance, and/or variant histology, we have shown that similar results can be obtained in selected high-risk patients compared with previously published data in low-risk patients (including some high-risk patients). ⁹

Further large-scale studies are needed to refine the criteria for selecting high-risk patients for endoscopic treatment, ideally on a prospective cohort of patients treated with endoscopy. In these future studies, an interesting debate will concern the ideal criteria that should be used to define success/failure of endoscopic management in high-risk UTUC. In our opinion, RNU should not be considered a primary endpoint, as delayed RNU still has the advantage of delaying RNU-related side effects, such as impaired renal function and cardiovascular events. ¹⁰ MFS and CSS appear to be more robust clinical criteria and should ideally be incorporated into future studies.

Finally, the benefit of endoscopic treatment is mainly associated with a lower risk of impaired renal function and long-term cardiovascular events. ¹¹ However, when discussing such alternative strategies with high-risk UTUC patients, the potential benefit of preserving renal function must be weighed against the disadvantages of endoscopic management, such as rigorous long-term monitoring and the burdensome nature of repeated procedures that can have a physical and psychologic impact on individuals. ¹²

Some limitations should be acknowledged. First, the main limitation of this study lies in its retrospective design. Second, a selection bias is recognized as endoscopic treatment was offered to selected high-risk patients without an invasive aspect on CT scan and in whom complete endoscopic treatment was considered feasible. Only a randomized trial will overcome this selection bias, which also affects all large retrospective studies conducted on RNU, as radically treated patients are more likely to harbor aggressive disease. ^3,4 Finally, all procedures were conducted in an expert tertiary center with its own endoscopic technique, and the results are not generalizable to all institutions.

Conclusions

In conclusion, we report the first large series of endoscopic treatment in patients with high-risk UTUC, arguing that promising oncologic outcomes can be achieved in properly selected candidates. We encourage multi-institutional collaborative work as a large cohort of high-risk patients treated endoscopically may allow subgroup analyses to define the best candidates.