Germline Variants Provide Insight Into Tumor Growth in Prostate Cancer

Paul C. Boutros, PhD

Paul C. Boutros, PhD

Select DNA variants leading to methylation dysregulation within the germline DNA of men with localized prostate cancer were found to be predictive of disease biology, suggesting the prognostic capacity of inherited elements of a patient’s genome, according to data published in Nature Medicine.^1,2

In the study, 1178 loci were found to be susceptible to DNA methylation that facilitated tumor growth, 17 of which showed tumor-specific RNA or protein level changes.

“Our work reveals a new landscape of mechanisms by which germline features influence prostate cancer behavior,” senior author, Paul C. Boutros, PhD, director of Cancer Data Science at the University of California, Los Angeles Jonsson Comprehensive Cancer Center, wrote in a statement to OncLive.

Although genome-wide association studies have identified germline variants that are linked with an increased risk of cancer, none have shown a robust association with prostate cancer proliferation.²

This served as the basis for the trial, which included 241 newly diagnosed patients with intermediate-risk localized prostate cancer. All patients had undergone germline whole-genome sequencing and methylation profiling prior to treatment with image-guided radiotherapy or surgery. Investigators supplemented these samples with 348 samples from The Cancer Genome Atlas.

Utilizing specialized computer sequencing software programs, the investigators analyzed the samples to find patterns in DNA, specifically if some patients with particular DNA variants had more or less methylation than those without such variants. The goal of the work was to identify locations where certain DNA variants made it easier or more difficult for tumors to control cancer genes through methylation. With these data, investigators could then determine which DNA variants had a higher likelihood of facilitating tumor growth.

Significant tumor-specific variants associated with microRNA expression are summarized according to the single nucleotide polymorphism, methylation probe, and gene identified, respectively:

• rs1225741, cg13351621, SYCP2L
• rs16934152, cg13558087, POLR1E
• rs2456274, cg08881796, VPS53
• rs2570972, cg08367326, AMIGO1
• rs3761188, cg09328228, PABPC1L
• rs3761188, cg15588266, PABPC1L
• rs3764509, cg14963724, CNDP2
• rs3807032, cg24330456, RNF39
• rs3807033, cg05563515, RNF39
• rs3807033, cg17322683, RNF39
• rs3807033, cg23793213, RNF39
• rs3849767, cg18264728, DAB2
• rs4147470, cg03997398, ABLIM3
• rs4147470, cg04669407, ABLIM3
• rs9261309, cg13918754, RNF39
• rs9261309, cg20249327, RNF39
• rs9295763, cg20249327, ELOVL2

Investigators selected 58 methylation probes based on their association with biochemical recurrence, defined by increasing prostate-specific antigen levels after primary therapy. Notably, not all methylation probes triggered tumor-specific changes in DNA methylation. Only 28% of the methylation probes that were associated with a higher likelihood of biochemical recurrence were targeted by changes in DNA methylation.

“The tumor-specific role of some germline variants in modulating tumor methylation is particularly interesting as these germline variants are present in nonmalignant tissue, but their regulatory role appears to only be triggered in a tumorigenic environment,” first author, Katie Houlahan, PhD candidate, Ontario Institute for Cancer Research, told OncLive.

Although these tumor-specific loci were not enriched in a particular pathway, 6 of 10 genes were more common in tumor versus normal tissue (false discovery rate [FDR] <.05). Moreover, in an exploratory analysis, investigators evaluated whether the 10 genes in the 17 tumor-specific variants were dysregulated at a protein level. VPS53 was found to be correlated with tumor loci (P = 6.95 x 10⁻¹²; Spearman’s |ρ| = −0.42) according to RNA expression (FDR = 8.22 x 10⁻³; ρ = −0.25) and protein levels (P = 4.27 x 10⁻²; ρ = −0.24), further validating the methods used to demonstrate inherited susceptibility.

“A large proportion of how prostate cancers evolve and present is influenced by a patient’s germline genome, said Boutros. “While much more work is needed to understand those influences, these data provide a lot of hope that robust biomarkers can be generated to guide improved screening, early detection of aggressive disease, and ultimately, be used to guide management decisions.”

In addition to these genetic variations, AKT1, which is a known prostate cancer driver, was found to be predictive of aggressive disease biology, and therein relapse after definitive local therapy, in the investigational (HR, 2.85; P= 5.8 x 10^-3) and validation cohorts of the trial (HR, 2.2; P= 1.7 x 10^-2).

“It’s unclear if these effects occur in other cancer types besides prostate. We would certainly expect them to, but that’s a key question of interest,” concluded Boutros. “It’s also unclear if the clinical correlations we observe are sufficiently strong enough to be useful in new biomarkers to aid in risk stratification. Additional validation studies will be needed to determine that.”

References

1. Clues from DNA could help predict growth of prostate cancer [news release]. Los Angeles, CA: UCLA Health. Published October 7, 2019. bit.ly/2oTwMt8. Accessed November 4, 2019.
2. Houlahan KE, Shiah YJ, Gusev A, et al. Genome-wide germline correlates of the epigenetic landscape of prostate cancer. [published online October 7, 2019]. Nat Med. doi: 10.1038/s41591-019-0579-z.