Clinical implications of a prostate cancer risk SNP profile in 2 treatment cohorts


Year:

Session type:

Chee Goh1, Edward Saunders1, Daniel Leongamornlert1, Tokhir Dadaev1, Malgorzata Tymrakiewicz1, Karen Thomas2, Elizabeth Selvadurai2, Ruth Woode-Amissah2, Nadiya Mahmud1, Elena Castro3, David Olmos3, Michelle Guy1, Koveela Govindasami1, Rosemary Wilkinson1, Emma Sawyer1, Ali Amin Al Olama4, Douglas Easton4, Zsofia Kote-Jarai1, Chris Parker1,2, Rosalind Eeles1,2
1The Institute of Cancer Research, London, UK, 2The Royal Marsden NHS Foundation Trust, London, UK, (CNIO) Spanish National Cancer Research Centre, 4The University of Cambridge, Cambridge, UK

Background

Genome-wide association studies (GWAS) have identified 77 single nucleotide polymorphisms (SNPs) associated with prostate cancer (PCa) risk. Currently their clinical utility remains undefined. We explore the prognostic role of these SNPs in 2 PCa treatment cohorts.

Method

DNA from patients undergoing active surveillance (AS) or androgen deprivation therapy (ADT) at the Royal Marsden Hospital was analysed. Risk SNPs were genotyped using Sequenom or Fluidigm platforms. The cumulative SNP risk scores for each patient were calculated by summing risk alleles for each of the loci using the weighted effect as estimated in previous studies (log-additive model). For the AS group, the risk scores were analysed against biopsy upgrade and time-to-treatment, to determine their prognostic value. For the ADT group, the risk scores were assessed against time-to-relapse, defined by either biochemical relapse or progression on repeat imaging.

Results

For the AS analyses, 391 patients' DNA was studied; 31% of those have since undergone treatment and 30% have histological upgrade on repeat biopsies. On univariate analysis, no significant relationship was found when risk scores were analysed against biopsy upgrade or time-to-treatment (p>0.05). There were also no significant differences in outcomes between men in the highest and lowest 25% of the genetic risk distribution.

For the ADT analyses, 567 patients' DNA was studied. The mean duration of response was 30 months. No significant associations were found when risk scores were analysed as a continuous variable against time-to-relapse, or when analysing differences between the highest and lowest 25% of the risk distribution.

Conclusion

PCa SNP risk scores have not been shown to be prognostic factors in either AS or ADT and so may be more useful for population risk stratification for screening. To our knowledge, this is the first study looking at the potential prognostic significance of the latest GWAS risk SNPs in PCa treatment cohorts.