Genetic Model Predicts Age of Onset for Aggressive Prostate Cancer

Tyler M. Seibert, MD, PhD

By combining data from genome-wide association studies (GWAS) with epidemiological surveys, investigators say they have developed a genetic tool that can predict age of onset for aggressive prostate cancer.

In the independent validation set, taken from the UK ProtecT trial, the hazard score calculated from 54 single nucleotide polymorphisms was a highly significant predictor of age at diagnosis of aggressive cancer (P <10⁻¹⁶). When men in the validation set with high scores (>98th centile) were compared with those with average scores (30th-70th centile), the risk was clearly greater for those with high scores (HR, 2.9; 95% CI, 2.4-3.4). The score was also predictive of any prostate cancer (HR, 2.5; 95% CI, 2.2-2.8; P <10⁻¹⁶) and very aggressive prostate cancer (HR, 3.0; 95% CI, 2.2-4.0; P <10⁻¹¹).

“The polygenic hazard score was calculated from 54 SNPs and proved to be a highly significant predictor of age at diagnosis of aggressive prostate cancer,” lead author Tyler M. Seibert, MD, PhD, chief resident physician in the Department of Radiation Medicine and Applied Sciences, UC San Diego School of Medicine, said in a press release.

“When men in the ProtecT dataset with a high polygenic hazard score were compared to those with average PHS, their risk of aggressive prostate cancer was at least 2.9 times greater. And when we account statistically for the effect of the GWAS having disproportionately high numbers of men with disease compared to the general population, we estimate that the risk defined by the polygenic hazard score is 4.6 times greater,” added Seibert et al.

The polygenic hazard score was developed previously as a parsimonious survival analysis model to predict the time to a given outcome. In this case, the outcome was age of onset for aggressive prostate cancer. The score is defined as the vector product of a patient’s genotype (X_i) for n selected single nucleotide polymorphisms and the corresponding parameter estimates (_i) from a Cox proportional hazards regression.

To develop the polygenic hazard score model, an international team of researchers led by the University of San Diego School of Medicine collected genotype and data on age from 21 studies from the PRACTICAL consortium. The data included 18,868 men with any prostate cancer, 10,635 with aggressive prostate cancer, 5406 with very aggressive disease, and 12,879 controls. All participants were of genotypic European ancestry. Age was taken either at diagnosis for patients or at last follow-up for the control group.

A total of 201,043 single nucleotide polymorphisms were available for analysis. Tumors in 4803 men could not be classified as aggressive or not because of incomplete data on staging, and these men were excluded from this analysis.

To verify whether the polygenic hazard score could accurately predict age of onset, investigators calculated the score for all patients in the validation set and tested it as the sole predictive variable in a Cox proportional hazards regression model for age of diagnosis. Patients in the validation set with a diagnosis of low risk disease (Gleason score ≤6, PSA concentration <10 ng/L, and stage T2N0M0 or lower) were censored at time of diagnosis because it could not be known if they would later receive a diagnosis of aggressive disease or at what age that might occur.

In the validation set, patients with polygenic hazard scores above the 50th centile accounted for 76% of cases of aggressive PCa, and the upper fifth accounted for 42%.

Investigators expected the polygenic hazard score to modulate the positive predictive value of PSA testing in the validation set. The positive predictive value of PSA was lower among patients with a low score and higher among patients with progressively higher scores. This pattern also held for the positive predictive value for any prostate cancer.

“The existing PSA test is useful, but it is not precise enough to be used indiscriminately on all men,” said Seibert.

Investigators knew the family history for 1405 patients and 4298 controls in the validation set. Family history alone was not predictive of age of onset of aggressive prostate cancer (HR, 1.1; 95% CI, 0.9-1.4; P = .37), though investigators noted a trend toward prediction for any prostate cancer (HR, 1.2; 1.0-1.3; P = .05). Inclusion of family history did not improve prediction over the polygenic hazard score alone for aggressive (P = .59) or any prostate cancer (P = .14). The score remained predictive when adjusted for family history.

Seibert TM, Fan CC, Wang Y, et al. Polygenic hazard score to guide screening for aggressive prostate cancer: development and validation in large scale cohorts [published online January 10, 2018]. BMJ doi: 10.1136/bmj.j5757.