Whole Genome Sampling Identifies Targetable Prostate Cancer Subgroups

By Peter M. Goodwin

CHICAGO—Clinically targetable mutations were identified and patient subgroups pinpointed by whole genome sampling (WGS) of castration-resistant metastatic prostate cancer (mPC) in a study reported at the 2018 ASCO Annual Meeting (Abstract 5014).

Lisanne F. van Dessel, MD, a medical oncologist at Erasmus Medical Center Cancer Institute in Rotterdam, The Netherlands, told Oncology Times that, while their research confirmed mPC was complex and unstable, they found that tumors having potentially actionable genomic features that could be targeted therapeutically were likely to be prognostic or predictive. This could help individualize therapies with existing targeted agents as well as develop new ones.

“We did find the ‘usual suspects’—namely the androgen receptors that are really well-known in prostate cancer. We [detected] a lot of alterations in the androgen receptors. But we also observed a distinct subgroup of patients with high tumor mutational burden associated with gene alterations in DNA repair genes and also associated with a specific [gene] signature for microsatellite instability (MSI) or mis-match DNA repair deficiency,” she explained.


Research Details

The data were derived from a cohort of 145 patients with castration-resistant mPC, each of whom gave a blood sample and had a tumor biopsy of the metastatic site from which fresh frozen samples were analyzed by WGS.

The sequencing data were checked for specific single nucleotide variants (SNV), small insertions and deletions (InDel), copy number alterations (CNA), and chromosomal rearrangements. The incidence of these features was compared to historical results reported in primary disease.

The study found that WGS of metastatic tissue from patients with mPC showed a clear evolution of aberrations compared to primary disease. There was increased frequency of alterations in specific driver genes confirming their role in disease dissemination and therapy failure.

The analysis was able to stratify patients into clinically relevant groups, including those with MSI, those sharing molecular features in common with mutated BRCA cancer susceptibility genes, and patients with phosphoinositide 3-kinase (PI3K) mutations. These groupings provided differential evidence for treatment decision-making that was distinct from existing evidence.

The study found that high SNV and InDel load was correlated with MSI mutational signatures in eight out of 153 tumors, and a distinct subgroup harboring “Signature Three” mutations was likely to be associated with BRCA deficiency in 26 out of 153 tumors.

The median number of interchromosomal rearrangements was 58 compared to 19 in primary disease. The study found a total of 131 gene alterations (including SNVs and CNAs) in either AR or TP53 genes— both important drivers in mPC.


Clinical Insights

When van Dessel was asked to explain what potential clinical insights the study added to existing knowledge—which already included specific gene testing for the targets of existing directed therapies for mPC—she said that with whole genome sequencing analysis they had a better overview of the whole genomic landscape. “So, we can also investigate structural variations—really large alterations in the genome—that you cannot identify with PCR-based targeted sequencing approaches.”

Signature analysis was more accurate if it was derived from the whole genome, she said. “You can then really identify the signatures, which can also be informative for targeted treatment, because we see that specific signatures are associated with microsatellite instability [that can identify] patients who could benefit from pembrolizumab, for instance.”

Van Dessel concluded that WGS made it possible to distinguish subgroups of patients who can benefit from specific targeted therapies. “We have a group with Signature Three, which is associated with BRCA alterations. And we also saw patients with this signature who did not have a BRCA alteration,” she said. Such patients could still benefit from PARP inhibitors, she suggested. “And you would not select those patients with the regular testing for those trials or treatments. So that’s something that can be done using whole genome sequencing analysis.

“The take-home message is that metastatic biopsies are feasible in castration-resistant prostate cancer patients and that whole genome sampling can identify extra patients for specific targetable treatments.”

Peter M. Goodwin is a contributing writer.