A Systematic Review and Meta-analysis of Clinicopathologic Factors Linked to Intravesical Recurrence After Radical Nephroureterectomy to Treat Upper Tract Urothelial Carcinoma

Abstract

Context

There is an ongoing debate about the factors that influence intravesical recurrence (IVR) after radical nephrouretectomy (RNU) to treat upper tract urothelial carcinoma (UTUC).

Objective

To assess significant predictors of IVR after RNU from a systematic review of the literature and meta-analysis.

Evidence acquisition

A computerized bibliographic search of the Medline, Embase, and Cochrane databases was performed for all reports that included detailed results of multivariate analyses on the predictors of IVR. According to the Preferred Reporting Items for Systematic Reviews and Meta-analysis (PRISMA) guidelines, we selected 18 retrospective studies that each included more than 100 patients treated exclusively with RNU between 2007 and 2014. Cumulative analyses of available hazard ratios (HRs) and their corresponding 95% confidence intervals were conducted using R software to assess the potential predictors of IVR.

Evidence synthesis

Among the 8275 patients included, 2402 (29%) were diagnosed with IVR within a median time of 22.2 mo (range 6.7–56.5). Patient-specific predictors were as follows: male gender (HR 1.37; p < 0.001), previous bladder cancer (HR 1.96; p < 0.001), and preoperative chronic kidney disease (HR 1.87; p = 0.002). Tumor-specific predictors were as follows: positive preoperative urinary cytology (HR 1.56; p < 0.001), ureteral location (HR 1.27; p < 0.001), multifocality (HR 1.61; p = 0.002), invasive pT stage (HR 1.38; p < 0.001), and necrosis (HR 2.17; p = 0.02). Treatment-specific predictors were as follows: a laparoscopic approach (HR 1.62; p = 0.003), extravesical bladder cuff removal (HR 1.22; p = 0.02), and positive surgical margins (HR 1.90; p = 0.004).

Conclusions

A meta-analysis of available data identified significant predictors of IVR that should be systematically assessed to propose a risk-adapted approach to adjuvant intravesical instillation of chemotherapy and cystoscopic surveillance after RNU.

Patient summary

In this report, we looked at the factors linked to intravesical recurrence after radical nephroureterectomy to treat upper urinary tract urothelial carcinoma. We identified patient-, tumor- and treatment-specific characteristics that should be systematically assessed to guide postoperative decision-making.

Introduction

The pathogenesis of intravesical recurrence (IVR) after radical nephroureterectomy (RNU) for upper tract urothelial carcinoma (UTUC) is still unclear. Subsequent bladder tumors could theoretically result from either implantation of a single transformed cell after descendant intraluminal seeding [1] or a pan-urothelial field defect [2], [3]. However, data to support a mixed monoclonal and oligoclonal origin of metachronous multifocal urothelial carcinoma have recently suggested that both mechanisms might be involved in the development of bladder cancer (BCa) following previous UTUC [4], [5], [6], [7]. Accordingly, two prospective randomized clinical trials have demonstrated that the clearance of any residual tumor cells, using a single early intravesical instillation of mitomycin C (MMC) or pirarubicin (THP) after RNU, decreased the risk of IVR [8], [9]. This strategy has now been advocated in the most recent European guidelines [10].

Nevertheless, there are huge discrepancies around the world regarding indiscriminate postoperative administration of chemotherapy into the bladder immediately after RNU. Clinical concerns have been raised with regard to the side effects of such a systematic strategy and, notably, the risk of painful extravasation into extraperitoneal tissues or potentially lethal intraperitoneal leakage, both of which are related to the hypothetically delayed healing of the bladder cuff removal area [8]. Therefore, accurate prediction of IVR for each patient might pinpoint those who are the best candidates for such an adjuvant local treatment. However, there are sparse data regarding predictors that could be used, not only to guide a risk-stratified approach to the adjuvant intravesical instillation of chemotherapy but also to adapt the frequency of cystoscopies during follow-up. Most studies attempting to identify the risk factors for IVR are limited by their single-center nature, small sample size, and heterogeneous population, including patients treated with either RNU or kidney-sparing surgery [11], [12], [13], [14], [15]. Although a nomogram has recently been developed to accurately assess the risk of IVR after RNU [16], external validation of this tool is still required before its routine use [17]. Therefore, our purpose was to assess significant predictors of IVR after RNU for UTUC from a systematic review of the literature and a meta-analysis of the available data.