Newly Identified Molecular Subtypes Could Guide Treatment Decisions in Prostate Cancer

Daniel E. Spratt, MD

Researchers have identified 3 distinct molecular subtypes of primary prostate cancer that correlate with distant metastasis-free survival (DMFS) and response to radiation therapy.

These subtypes, along with the standardization of genetic testing, will pave the way toward a more effective, personalized approach to prostate cancer treatment, said lead study author Daniel E. Spratt, MD, who presented the findings at the 2016 ASTRO Annual Meeting.¹

Current methods of identifying a patient’s risk for recurrence are lacking, he explained.

“A big challenge for an oncologist in prostate cancer is that the risk groups we currently use to guide how we treat our patients are made up of features that are relatively insensitive, such as a blood test, prostate-specific antigen (PSA), or Gleason score,” Spratt, chief of the Genitourinary Radiotherapy Program at the University of Michigan Comprehensive Cancer Center, said in an interview with OncLive.

“These features don’t really tell us exactly what a patient’s disease looks like genetically. Because of that, we have to sort of look at these phenotypes—the way all these features group together—to guide the treatment for our patients, rather than actually genetically look at how aggressive these tumors are or how responsive they would be to our therapies.”

The subtypes were identified in an analysis of 4236 primary prostate cancer samples from 9 independent cohorts. This was the largest study of its kind to date.

In an effort to create intrinsic subtypes specific to the cancer itself rather than the associated surrounding tissue, data were refined to remove nontumor genes from the dataset. Through K-median clustering analyses, researchers identified 3 molecular subtypes of prostate cancer that could be characterized through a profile of 100 distinct genes, named the Prostate Cancer 100 (PC100) by study investigators.

Researchers validated the subtypes across 6 additional retrospective cohorts, representing a variety of RNA sequencing platforms and tissue storage methods, as well as 2 prospective cohorts comprising 2610 patients. The intrinsic subtypes were associated with androgen receptor activity, expression of the ERG oncogene and other known drivers of prostate tumor growth and progression, but researchers did not find a link from mutations or genetic rearrangements to the subtypes.

The 10-year rates of DMFS varied significantly across the 3 subtypes, with a 73.6% rate of DMFS in subtype A, 64.4% in subtype B, and 57.1% in subtype C (B vs A: cox HR, 1.31, P =.02; C vs A: HR, 1.65, P =.0001). The subtypes were independently associated with DMFS after adjusting for clinicopathologic variables (B vs A: cox HR, 1.31, P = .026; C vs A: HR, 1.33, P =.024). In addition, subtype C was significantly associated with response to postoperative radiotherapy (P = .0016).

These findings could help radiation oncologists better make decisions regarding radiotherapy for patients, said Spratt.

“A frustration for a lot of clinicians, including myself, is that when we look at a patient that we would—for example—term an intermediate-risk patient, we treat them the same as any other risk patient,” he said. “We treat them all with radiation therapy, although some may use different types of radiation and some might add hormonal therapy. We don’t truly know how aggressive their cancer is, nor do we know if they will respond favorably to our therapies. That is a difficulty for any clinician that we group these patients by clinical and pathologic features—not genetic features.”

Utilizing these subtypes to make treatment decisions would require more oncologists to take advantage of genetic testing. This is already happening in several practices, said Spratt. Additionally, there are already a variety of tests that are commercially available and recommended by NCCN guidelines for use in prostate cancer.

“We are already in an era where more and more patients are going to get genetically tested,” said Spratt. “In breast cancer, they already have a test that helps predict risk of recurrence and response to chemotherapy. As we adopt this into prostate cancer, it is really going to help us personalize treatment for all men with the disease.”

Spratt DE, Zhao SG, Chang SL, et al. Identification and validation of intrinsic subtypes of prostate cancer. Int J Rad Onc. 2016;96(2S): S3—S4.