Subtypes of urothelial cancer identified through gene expression profiling predict which patients are most likely to respond to neoadjuvant cisplatin-based chemotherapy, according to research from The University of Texas MD Anderson Cancer Center. These findings will pave the way for a more personalized approach to treatment for urothelial cancers, including difficult-to-treat muscle-invasive bladder cancer.
The work was led by Arlene Siefker-Radtke, M.D., an associate professor in the Department of Genitourinary Medical Oncology. She said, “For decades, we’ve been treating bladder cancer as if it’s all one disease. And until recently, we didn’t have the techniques available to try to predict responders to specific therapies. Using gene expression profiling, we can start to understand the biology of different bladder cancers and predict which tumors will respond to specific therapies.”
Bladder cancer subtypes
In recent years, researchers from The Cancer Genome Atlas, MD Anderson, and other groups have shown through gene expression profiling that urothelial cancers can be divided into three subtypes: basal, regular luminal, and p53-like, which is a distinct subset of the luminal subtype. Basal tumors have a stem cell–like appearance and proliferate rapidly. Regular luminal tumors share features with the umbrella cells of the bladder, tend to have enrichment for mutations in the gene encoding fibroblast growth factor receptor 3 (FGFR3) on gene set enrichment analysis, and have an intermediate rate of proliferation. p53-like tumors, which account for approximately half of luminal tumors, are characterized by a gene expression signature resembling that of wild-type p53, infiltration with stromal fibroblasts, and a slow proliferation rate.
Dr. Siefker-Radtke and her colleagues Woonyoung Choi, Ph.D., an assistant professor in the Department of Urology, and David McConkey, Ph.D., formerly a professor in the Department of Urology at MD Anderson and now director of the Johns Hopkins Greenberg Bladder Cancer Institute, decided to test whether these three subtypes predicted survival after neoadjuvant chemotherapy. Specifically, the researchers looked at tumor specimens from patients with muscle-invasive or other high-risk urothelial cancers who were treated at MD Anderson in a phase II trial of neo-adjuvant dose-dense methotrexate, vinblastine, doxorubicin, and cisplatin (MVAC) plus bevacizumab followed by cystectomy. Cisplatin-based neoadjuvant chemotherapy is offered to many patients with muscle-invasive urothelial cancer but proves beneficial in only 30%–40% of treated patients; the ability to predict up front which patients are likely to respond to such therapy would help oncologists tailor treatment.
Thirty-eight patients in the phase II trial had bladder specimens available for gene expression profiling. Of these, 16 patients had basal, 11 had regular luminal, and 16 had p53-like tumors. Despite the small sample size, the results clearly showed that patients with basal tumors had a higher 5-year overall survival rate (91%) than did patients with regular luminal tumors (73%) or p53-like tumors (36%; P = .015). The survival advantage of patients with basal tumors remained significant in a multivariable analysis that included age at clinical trial registration and the presence of lymphovascular invasion. An additional finding of interest was that bone metastases within 2 years were observed in nine of the 16 patients with p53-like tumors but in no patients with the other subtypes.
These findings suggest that information about urothelial cancer subtypes could be used to guide treatment decisions. Basal tumors, despite their aggressive features, were responsive to MVAC; therefore, patients with basal tumors could be good candidates for neoadjuvant cisplatin-based chemotherapy. In contrast, the observation of bone metastases in more than half of the patients with p53-like tumors suggests that patients with p53-like tumors should be treated with surgery first and might be good candidates for adjuvant treatment with agents targeting stroma or bone.
The predictive power of the urothelial cancer subtypes was then confirmed through gene expression analysis and survival analysis in a separate group of 49 patients who had been treated with MVAC in an earlier clinical trial. Results from these analyses confirmed the survival advantage of patients with basal tumors: the 5-year overall survival rates were 77%, 56%, and 56% for patients with basal, regular luminal, and p53-like tumors, respectively (P = .02).
Ongoing clinical trials
Researchers are now trying to learn whether urothelial cancer subtypes predict responses to targeted anticancer drugs. Dr. Siefker-Radtke is currently leading two trials that address this question. Both are open to patients with metastatic or surgically unresectable urothelial cancers already treated with chemotherapy and/or immunotherapy.
One trial (No. 2015-0112) is testing the safety and efficacy of two different doses of a pan-FGFR tyrosine kinase inhibitor in patients who have urothelial tumors with genomic alterations in FGFR3. This is one of several trials worldwide testing FGFR3 inhibitors in urothelial cancer. A significant proportion of both regular and p53-like luminal bladder cancers have FGFR3 mutations.
In another trial (No. 2014-0661), patients with metastatic, unresectable urothelial cancer receive the proteasome inhibitor ixazomib in combination with gemcitabine and doxorubicin. In pre-clinical studies, ixazomib increased the sensitivity of bladder cancer cells to chemotherapy. Dr. Siefker-Radtke and her colleagues plan to perform gene expression profiling of the patients enrolled in this study to find out if certain subtypes of urothelial cancer are more sensitive than others are to ixazomib. Some data suggest that ixazomib inhibits angiogenesis; thus, this drug may be more effective against basal urothelial cancer, which is highly proliferative, than against other subtypes.
At present, gene expression profiling is not the standard of care for patients with urothelial cancer. However, Drs. Siefker-Radtke, McConkey, and Choi are working with several private companies to further investigate the ability to predict response to chemotherapy and targeted agents. The researchers also hope to perform gene expression profiling on more patients with urothelial cancer who are enrolled in clinical trials. The resulting knowledge of the underlying tumor biology would allow rational development of therapies and combinations of therapies to target specific types of tumors.
“Bladder cancer is not just one disease,” Dr. Siefker-Radtke said. “We’re heading toward a more personalized approach to the treatment of our bladder cancer patients.”
For more information, contact Dr. Arlene Siefker-Radtke at 713-792-2830.
McConkey DJ, Choi W, Shen Y, et al. A prognostic gene expression signature in the molecular classification of chemotherapy-naïve urothelial cancer is predictive of clinical outcomes from neoadjuvant chemotherapy: a phase 2 trial of dose-dense methotrexate, vinblastine, doxorubicin, and cisplatin with bevacizumab in urothelial cancer. Eur Urol. 2016;69: 855–862.
OncoLog, November-December 2016, Volume 61, Issue 11-12