Biomarker May Predict Prostate Cancers Requiring Treatment

Not all early-stage prostate cancer diagnoses are alike. While some patients have aggressive tumors, others have slow-growing, low Gleason score tumors that may not require treatment, but instead can be monitored with regular clinical evaluations. But distinguishing between prostate cancers that require treatment and those that do not is still a major challenge.

Researchers at Columbia University in New York City have now identified a 3-gene signature that could indicate whether a particular early-stage prostate cancer is indolent. The test relies on a tissue sample, and along with a prostate-specific antigen (PSA) test and a histology assessment, could help clinicians make an accurate diagnosis. The early results, including a blinded retrospective analysis of 43 patients, show that the signature can accurately predict which patients with low-risk disease would develop metastatic prostate cancer and which patients would not progress. The study is published in Science Translational Medicine.

“These types of markers will, for the first time, give us the opportunity to measure biological features of cancer in the same patient, with multiple biopsies spread out over many years,” said Eric Klein, MD, chairman, Glickman Urological and Kidney Institute at the Cleveland Clinic in Ohio.

Cory Abate-Shen, PhD, professor of urological oncology at Columbia University; Andrea Califano, PhD, professor of systems biology at Columbia University; and colleagues used a computational approach that identified three genes-FGFR1, PMP22, and CDKN1A-all associated with aging, that could accurately predict outcomes of low-risk, low Gleason score prostate tumors. Protein and mRNA levels of all three genes were high in those patients who had non-aggressive, indolent disease and low in those who had aggressive tumors.

Clinicians still rely on the Gleason score, a histology and pathology evaluation that does not incorporate any molecular assessment. Those patients with a Gleason score of 8 or higher are candidates for immediate treatment, but whether men with a score of 6 or 7 require treatment is difficult to assess-no test exists to identify the small percentage of patients who have early-stage prostate cancer that is more likely to metastasize.

The 3-gene signature was validated using an independent prostate cancer cohort. According to the study authors, the signature was prognostic and improved prognosis compared with the use of PSA and clinical assessment.

“We would predict that the test would be beneficial for patients with low Gleason score prostate tumors,” said Abate-Shen. “These patients are now typically monitored on active surveillance protocols, and the patients get a biopsy periodically. The test would be conducted on the biopsy.”

Rather than focusing on the entire genome, the researchers focused on 377 genes involved in aging, predicting that genes involved in aging and senescence are critical for tumor suppression. Cellular senescence is known to play a role in tumor suppression and is associated with benign prostate tumors both in the clinic and in mouse models, according to the researchers. Using a computational analysis called gene set enrichment analysis (GSEA), they narrowed the long gene list to 19 genes, and then to a set of 3 genes that could identify indolent tumors.

“To focus on senescence genes is intellectually interesting,” said Klein. “There is already a body of work supporting the role of these genes in prostate cancer, but to my knowledge no one has looked at them in early-stage disease before.”

Forty-three patients, who had been under active surveillance for 10 years at Columbia University Medical School, were used for the blinded retrospective analysis to assess the predictive value of the gene signature. Each patient had been diagnosed with low-risk prostate cancer, with a Gleason score of 6 or less. The test was correctly able to identify all 14 patients who eventually developed advanced prostate cancer.

CDKN1A has been shown to be linked to senescence and to regulate the cell cycle. Previous studies have shown that downregulation of the gene is linked to cancer progression. The correlation of FGFR1 (fibroblast growth factor receptor 1) with indolent tumors was surprising, as fibroblast growth factors have been shown to play a role in prostate cancer development. But, as the authors highlight in their discussion, FGFR1 signaling in prostate cancer is likely complex. The third gene in the signature, PMP22, encodes a glycoprotein expressed in neurons and has not been previously associated with prostate cancer.

This 3-gene signature is different from previously identified biomarkers, which have largely focused on advanced tumors. The potential biomarker test could complement other approaches in development, such as urine or blood tests, according to the authors.

A trial to validate the genetic signature is underway at Columbia University, and a national trial is being planned.

“It is really important to find novel ways to help to define early-stage tumors that may or may not progress to aggressive disease,” said Abate-Shen. “This will ultimately really help to minimize overtreatment, while capitalizing on the benefits of cancer screening.”

Other genomic approaches to distinguish indolent and aggressive disease are also underway. The first-generation expression-based tests, including Oncotype DX prostate and Prolaris, can facilitate clinical decisions based on the molecular characteristics of a prostate tumor. Both the available tests and the new ones “promise to reduce overtreatment and help men make the right decisions based on biology rather than uncertainty,” said Klein.