Skip to main content
Premium Trial:

Request an Annual Quote

Transcriptional Signature Marks Advanced Prostate Cancer Subtype Arising From Treatment

NEW YORK (GenomeWeb) – A University of California, San Francisco-led team has characterized genomic, transcriptomic, and clinical features of a type of advanced small-cell neuroendocrine prostate cancer that arises after treatment with androgen receptor (AR)-targeted therapy.

"An understanding of the biology of this important mechanism of resistance is essential to our developing novel therapeutics designed to prevent the development of this lethal prostate cancer subtype, or, once developed, to effectively treat it," senior author Eric Small, hematology and oncology chief at UCSF, said in a statement.

Using a combination of targeted DNA and RNA sequencing, the researchers profiled formalin-fixed, paraffin-embedded samples from 202 consecutive cases of metastatic, castration-resistant prostate cancer, focusing on the genetic alterations, transcriptional signatures, and clinical characteristics found in individuals with treatment-emergent small-cell neuroendocrine prostate cancer (t-SCNC). As they reported in the Journal of Clinical Oncology, they found that pre-treated castration-resistant metastatic prostate cancer patients with t-SCNC tumors had shorter-than-usual overall survival times.

"We want to know why prostate cancer becomes resistant, and we believe the emergence of t-SCNC is one important mechanism through which they evolve and evade treatment," UCSF researcher Rahul Aggarwal, the study's first author, said in a statement.

The team noted that roughly three-quarters of the t-SCNC prostate cancer tumors expressed AR, while 67 percent harbored AR amplifications. The t-SCNC tumor transcriptomes typically clustered into a small-cell-like group, the group reported, leading to a transcriptional signature for this advanced prostate tumor type that was subsequently validated in an external tumor set.

From these and other findings, the authors concluded that "[t]ranscriptional profiling facilitates the identification of t-SCNC and novel therapeutic targets."

Whereas metastases from other cancer types tend to be profiled by biopsies, the team noted, that approach has traditionally been less common for advanced prostate cancers, since many of these metastatic tumors arise in the bone, which is relatively tricky to test.

"Obtaining tumor biopsies in metastatic cancer has not in the past been the standard of care," Aggarwal said, noting that "it is being done more often, in part to look for neuroendocrine tumor cells, but more generally to get an idea for what mutations are driving cancer growth."

To take a look at the frequency of t-SCNC and the features characterizing them, Small, Aggarwal, and their colleagues brought together available baseline and progressed tumor samples from 202 metastatic castration-resistant prostate patients enrolled sequentially between late 2012 and April 2016.

Nearly three-quarters of those advanced prostate cancer patients experienced disease progression on hormone or anti-androgen therapies such as abiraterone or enzalutamide, the researchers reported, including 20 tumors with pure t-SCNC histology and seven with mixed t-SCNC histology. They had access to corresponding biopsies for 160 metastatic tumors found in bone, liver, lymph node, or soft tissue, representing 79 percent of cases.

The authors noted that 27 of the 160 previously treated prostate cancers had features coinciding with t-SCNC. The alterations identified in these tumors were more or less mutually exclusive with the type of DNA repair mutations found in other advanced prostate cancers, they wrote, suggesting that t-SCNC "may be a distinct subset of [metastatic castration-resistant prostate cancer]."

From RNA-seq data for 119 baseline or progression tumor samples, meanwhile, the team tracked down a dozen transcriptome clusters for the advanced cancers. One of the clusters was enriched for t-SCNC tumors, with differential expression at 61 genes, including enhanced expression of the pancreatic neuroendocrine marker and transcription factor PDX1 and muted RB1 regulatory gene expression.

These insights made it possible to put together an expression signature that correctly classified the majority of t-SCNC or adenocarcinoma tumors, respectively, in additional advanced prostate cancer sets or in primary prostate cancers from the Cancer Genome Atlas Project. 

"Application of the t-SCNC expression signature to other external data sets, as they become available, will provide additional clarity regarding the incidence of t-SCNC," the authors noted.