Skip to main content
Premium Trial:

Request an Annual Quote

Bladder Cancer Response to Immunotherapy Informed by Expression Profiles

NEW YORK (GenomeWeb) – New research suggests that stromal cells in the tumor microenvironment may interfere with immune system's efforts to combat bladder cancer in patients treated with anti-PD-1 immunotherapy.

Using urothelial cancer data from the Cancer Genome Atlas project, patient xenografts models, and gene expression data for metastatic urothelial cancer patients treated with the PD-1 inhibitor nivolumab, researchers from the Icahn School of Medicine at Mount Sinai, the Mount Sinai spinout Sema4, and elsewhere found that the presence of infiltrating T-cells — the mark of 'hot' tumors — seemed to bode well for tumor response to the PD-1 inhibitor nivolumab, while a related epithelial-mesenchymal transition (EMT) expression signature did not.

Instead, the team reported today in Nature Communications, a rise in the EMT signature and expression of non-hematopoietic stromal cell genes dialed down the benefits of T-cell infiltration — potentially through immune exclusion from the tumor — leading to poorer response to nivolumab, and survival times in patients with chemotherapy-resistant metastatic urothelial cancer.

"Some bladder cancers may not respond to immunotherapy, even though the body has developed an immune response against them, because the T-cells are prevented from reaching the tumor by stromal cells that create an inhospitable 'neighborhood,'" co-corresponding author Matthew Galsky, director of genitourinary medical oncology at Mount Sinai's Tisch Cancer Institute, said in a statement.

Up to one-quarter of individuals with advanced, chemotherapy-resistant urothelial cancer have shown responses to immune checkpoint drugs targeting PD-1 or PD-L1 in past clinical trials, the team explained. But questions remain about strategies for finding the metastatic urothelial cancer cases most apt to respond to the treatment.

In the case of EMT, which has been previously associated with T-cell infiltration, the authors noted, "some studies have suggested that patients with tumors with higher EMT-related gene expression should be more likely to benefit from immune checkpoint blockade, whereas others have linked EMT-related gene expression with immunotherapy resistance."

When they analyzed gene expression profiles in more than 400 urothelial cancer samples from TCGA, the researchers narrowed in on 144 genes with enhanced expression in tumor-infiltrating T-cells — a signature they used to find other genes and pathways showing altered expression when T-cell infiltration was strong.

Although the EMT signature tended to coincide with infiltrating T-cell activity, though, the team's tumor purity analyses indicated that the EMT-related expression signal often stems from stromal cells and not from epithelial cells in the tumor itself. Moreover, the T-cell infiltration and EMT expression features appeared to have distinct ties to patient outcomes.

From the TCGA data and in the researchers' subsequent analyses of RNA sequence data from five urothelial cancer xenografts from mice, the EMT signature seemed to reflect stromal cell activity — findings that were shored up when they incorporated gene expression profiles for 270 individuals from the CheckMate 275 study with platinum-resistant metastatic urothelial cancer.

In the latter cohort, the team saw higher objective response rates, longer progression-free survival times, and enhanced overall survival in the cases with greater CD8 T-cell infiltration based on gene expression or immunohistochemistry. On the other hand, the EMT-stromal signature, though not independently associated with such outcomes, was linked to poorer prognoses in tumors with pronounced T-cell infiltration.

"Together, our findings suggest a stroma-mediated source of immune resistance in urothelial cancer and provide rationale for co-targeting PD-1 and stromal elements," Galsky and his colleagues wrote, noting that "in T-cell infiltrated tumors with increased EMT/stroma-related gene expression, T cells may be spatially separate from cancer cells."