Predicting Response to Intravesical Therapy in Non–muscle-invasive Bladder Cancer

Abstract

Context

The ability to predict response to intravesical therapy (IVT) following transurethral resection in non–muscle-invasive bladder cancer holds important prognostic information. However, few predictive tools are available to guide urologists.

Objective

We reviewed the most recent studies investigating the predictors of response to IVT.

Evidence acquisition

A literature search was conducted using PubMed database from January 1, 2013 to April 1, 2018 following the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) criteria. For our search strategy, we used the combination of the MeSH terms of "Administration, Intravesical" and "Urinary Bladder Neoplasms" with any of the following words: "Biomarkers," "Predictive Value of Tests," "response," "recurrence," and "progression." We limited our search to the English language.

Evidence synthesis

Risk stratification models utilizing clinicopathological features are the most cost-effective and widely used tools currently available to predict response to IVT. Additionally, urinary fluorescence in situ hybridization testing and urinary cytokine-based nomograms (Cytokine Panel for Response to Intravesical Therapy) may enhance predictive ability. Protein-based biomarkers have been associated with predicting recurrence. Several gene-based biomarkers quantifying mutations in DNA damage repair genes may have predictive ability. However, genomic data are relatively new and lack validation.

Conclusions

Clinicopathological criteria remain the most widely utilized tool for predicting IVT response. Further research to validate protein- and genomic-based biomarkers are needed before adoption in clinical practice.

Patient summary

We reviewed contemporary studies that investigated how to predict response to medication instilled in the bladder (intravesical therapy) for bladder cancer. We found that most predictive tools use clinical data, such as tumor stage and grade, to determine the outcome. Newer biological (gene, protein, cytokines) marker tests are being studied. We concluded that the combination of clinical data with levels of certain experimental markers (fluorescence in situ hybridization test or urinary cytokines) may improve predictive ability. Genetic testing methods may also yield additional predictive markers in the future, but this needs more validation.