SAN FRANCISCO — Neuroendocrine prostate cancer (NEPC) usually occurs as a late manifestation of prostate cancer, and can be associated with resistance to androgen-deprivation therapy (ADT).

Researchers are hopeful that a new gene-expression signature will help identify patients who have developed NEPC, because it has been used to show significant differences in metastasis-free survival between patients treated with ADT and those not treated with ADT.

The findings were presented during a poster session here at the Genitourinary Cancers Symposium 2016.

"With our signature, patients who got high scores will have a higher probability of developing metastasis if they are treated with hormone therapy," said researcher Hussam Al-Deen Ashab, MSc, a bioinformatician on the research and development team at GenomeDX Biosciences, the developer of the assay.

"Their risk was not very different from patients who didn't get treated because, according to the signature, they are resistant to therapy," he said.

NEPC can be de novo (small-cell prostate cancer), but more commonly arises after exposure to ADT, the researchers explain. It represents about 25% of late-stage disease, and prognosis is poor; most patients survive for less than 1 year after diagnosis.

"Some of these patients will develop resistance to treatment and some will develop a phenotype that will lead to metastasis and more aggressive disease if they are treated just with hormone therapy," Ashab told Medscape Medical News. "What we are doing is trying to find the subgroup of patients who will fail treatment. This will help physicians make more optimal therapeutic decisions."

This study was actually conducted on prostate adenocarcinomas, but known NEPC genetic mutations were used, as were androgen-receptor signaling and other mutations, said senior researcher R. Jeffrey Karnes, MD, associate professor of urology at the Mayo Clinic in Rochester, Minnesota. However, he explained, "the best were from the NEPC."

"NEPCs are well known to be resistant to conventional ADT," Dr Karnes told Medscape Medical News.

In their study, Dr Ashab and his colleagues obtained expression profiles for 1023 patients with prostate cancer — 243 who had received ADT and 780 who had not. Of this cohort, 529 were used as the training set and 494 were used as the validation set.

The team conducted a literature review of ADT resistance and neuroendocrine genes, and selected 1557 candidate genes. This was then filtered using logistic regression, which ultimately produced a 52-gene ADT-resistance signature.

In two validation sets, the signature was predictive of metastasis in patients who received adjuvant ADT. The 10-year metastasis-free survival c-index was higher in patients treated with ADT than in those we were not (0.69 vs 0.45).

Results were similar in a separate group of patients not treated with ADT until after they developed metastasis, but in these patients, the signature was not prognostic (c-index, 0.53).

In patients treated with ADT, 10-year metastasis-free survival was better when the signature score was low than when it was high (87% vs 70%; P < .001).

In the 41 men who received ADT after metastatic onset and who developed castrate-resistant prostate cancer, median time to treatment failure was shorter in those with a high ADT-resistance signature score than in those with a low score (1 vs 2 years; P = .07).

The resistance-gene signature is not yet ready for prime time; even though it was validated in two separate sets, it needs validation in a randomized trial. The next step is to look at higher-risk prostate cancers in men who do not undergo surgery but who receive radiation, often concomitant with ADT, Dr Karnes explained.

It is not known at this point whether the signature can predict failure on a biopsy. "Work could also be extended to newly diagnosed metastatic disease to identify whether to give ADT or something else, like ADT plus chemotherapy, but we have no data on this," he said.

The fact that prostate cancer is a heterogenous disease, both clinically and at the molecular level, was pointed out by Himisha Beltran, MD, assistant professor of medicine and urology at Weill Medical College of Cornell University in New York City.

"This has potential biologic implications in elucidating innate versus acquired resistance to primary ADT," Dr Beltran told Medscape Medial News. "If validated, this approach could aid the development of biomarkers to help select the most effective adjuvant therapy for patients and/or the design of trials looking at alternative targeted or cotargeting adjuvant strategies for nonresponders."

Dr Karnes has disclosed no relevant financial relationships. Mr Ashab is employed by GenomeDx. Several coauthors report relationships with GenomeDx and other companies, as noted in the abstract.

Genitourinary Cancers Symposium (GUCS) 2016: Abstract 106. Presented January 7, 2016.

Bobby