Molecular profile of invasive bladder cancer resembles that of breast cancer

MD Anderson News Release February 18, 2014

Similarities point to possible chemotherapy guidance and targeted therapy for high-grade disease

MD Anderson News Release 02/18/14

Researchers who took a fresh look at muscle-invasive bladder cancer through the lens of gene expression discovered that it looks remarkably like breast cancer. This resemblance has important implications for treating the most lethal form of bladder cancer.

Scientists at The University of Texas MD Anderson Cancer Center reported in the February edition of Cancer Cell that the gene expression profiles of advanced bladder cancer fall into three molecular categories that closely resemble three of the four major subtypes of breast cancer.

“Several of our findings have immediate potential impact on how we address muscle-invasive bladder cancer with chemotherapy,” said study senior author David McConkey, Ph.D., professor of Urology. 

“There are no targeted therapies for this high-grade cancer now, so a future implication of these findings is developing new, better approaches for treating our patients,” McConkey said. Characterization of breast cancer is more advanced, with targeted approaches available for three subtypes and chemotherapy advised for the fourth.

Muscle-invasive disease makes up about 30 percent of bladder cancer cases but causes the vast majority of deaths.  It’s treated with chemotherapy, surgery and radiation.  About 15,000 Americans will die of the disease this year.

Guidance for chemotherapy, potential targeted drugs

“We know that cisplatin-based chemotherapy combinations work for about 30-40 percent of cases, but there’s no way to identify patients in advance who are likely to benefit,” McConkey said.

McConkey and colleagues identified a basal subtype of invasive bladder cancer that’s aggressive but vulnerable to chemotherapy and a p53-like luminal subtype that’s highly resistant to chemotherapy. These observations could lead to pre-treatment tumor analysis that guides the chemotherapy decision.

The third subtype of high-grade bladder cancer (luminal, similar to luminal B breast cancers) may be vulnerable to targeted therapies used in those subtypes of breast cancer, including estrogen receptor blockers.

• The basal subtype of muscle-invasive bladder cancer expresses genes that are biomarkers for basal breast cancer (CD44, KRT5, KRT6 and CDH3) and indicate the presence of bladder cancer stem cells and other treatment-resistant features.  Like its breast cancer counterpart, the authors noted, it’s a biologically aggressive cancer, causing shorter overall survival for patients treated with surgery alone. But it’s also sensitive to chemotherapy, so patients with basal tumors should be treated aggressively with frontline cisplatin-based regimens.  In breast cancer, basal cancers are known as triple-negative disease, because they lack drug targets for hormone receptors or the growth factor protein HER2. Some of them are also highly sensitive to chemotherapy.
• The luminal subtype expresses biomarkers shared by the luminal A and B subtypes of breast cancer (CD24, FOXA1, GATA3, ERBB2).  These markers are involved in breast cancers driven by the hormones estrogen and progesterone, raising the possibility that drugs hitting these targets may be effective for some bladder cancer patients.  The tumors are also enriched with activating mutations in a growth factor receptor (FGFR3) that can be targeted by drugs (FGFR inhibitors) in clinical development.  These patients have a better prognosis than those with basal subtype.
• The p53-like subtype consists of tumors that have gene expression patterns that are similar to those stimulated by p53, which is an important tumor-suppressor and is often mutated, deleted or otherwise inactivated in cancer.  However, it doesn’t appear that p53 itself is responsible for these gene expression patterns.  Surprisingly, the team found that these p53-like bladder cancer tumors resist chemotherapy.  This subtype of bladder cancer closely resembles luminal A breast cancers, which are estrogen-receptor positive but proliferate more slowly than luminal B breast cancer.

Cell dormancy caused by p53 a potential factor in resistance

The p53-like subtype finding was surprising because a central tenet of cancer biology is that p53 expression is required for programmed cell death (apoptosis) of cancer cells in response to DNA damage.

However, McConkey said, there’s ample support for a treatment-resistance effect both in the team’s research and discoveries by others in breast cancer.

The p53-like effect held up across multiple cohorts of bladder cancer tumors studied and was implicated in both immediate and acquired chemotherapy resistance.  While p53’s role in programmed cell death is well-known, it also can simply arrest cell growth and division, essentially putting it to sleep.

“These dormant cells evade chemotherapy, which preferentially kills dividing cells,” McConkey said.  They presumably awaken later to spark recurrence or metastasis. A recent clinical trial of presurgical chemotherapy for breast cancer called I-SPY1 found that tumors with normal p53 expression signatures also responded poorly to presurgical chemotherapy.

Two years ago, Guillermina Lozano, Ph.D., chair and professor of Genetics at MD Anderson, found that p53 expression in a mouse model of breast cancer caused senescence and chemotherapy resistance, McConkey said.

The researchers suggest that gauging p53 pathway gene expression, rather than the presence of p53 genetic mutations, would more accurately predict chemotherapy resistance.

Next stop, phase II prospective clinical trial

A multi-institutional prospective phase II clinical trial for muscle-invasive bladder cancer led by the Southwest Oncology Group (SWOG, S1314, also called the “CoXEN” trial) opens soon that will allow an even more rigorous assessment of the subtypes’ clinical value to guide therapy, McConkey said.

The team is developing streamlined methods for identifying these subtypes and using them to subtype MD Anderson bladder cancer patients.

The researchers in this project first analyzed 73 flash-frozen bladder cancer tumor samples from MD Anderson. They then validated the initial findings in a more conventional set of 57 formalin-fixed, paraffin-embedded samples.  The findings were confirmed in publicly available gene expression profiling data sets.  They also performed subtype analyses on tumors collected in a phase II clinical trial performed by researchers at Fox Chase Cancer Center and Thomas Jefferson University Hospital in Philadelphia.

Co-authors with McConkey are lead author Woonyoung Choi, Ph.D., Sima Porten, M.D., Daniel Willis, M.D., Beat Roth, M.D., Tiewei Cheng, Mai Tran, I-Ling Lee, Jonathan Melquist, M.D.,  Shanna Pretzsch, and Colin Dinney, M.D., all of Urology; Arlene Siefker-Radtke, M.D., of Genitourinary Medical Oncology; Jolanta Bondaruk, Ph.D., Tadeusz Majewski, Ph.D., Shizhen Zhang, Ph.D., and Bogdan Czerniak, M.D., Ph.D., of Pathology; Keith Baggerly, Ph.D., of Bioinformatics; Seungchan Kim, Ph.D., of the Translational Genomics Research Institute in Phoenix; Elizabeth Plimack, M.D., of Fox Chase Cancer Center in Philadelphia and Jean Hoffman-Censits, M.D., of Thomas Jefferson University Hospital, Philadelphia.

Cheng and Tran are graduate students in The University of Texas Graduate School of Biomedical Sciences at Houston, which is jointly operated by MD Anderson and The University of Texas Health Science Center at Houston.

This research was funded by a grant from the Dexter F. and Dorothy H. Baker Foundation, the MD Anderson Bladder Specialized Program in Research Excellence (SPORE) grant from the National Cancer Institute of the National Institutes of Health; and MD Anderson’s Cancer Center Support Grant from the NCI.