Prostate cancer risk stratification improvement across multiple ancestries with new polygenic hazard score
Minh-Phuong Huynh-Le,Roshan Karunamuni,Chun Chieh Fan,Lui Asona,Wesley K Thompson,Maria Elena Martinez,Rosalind A Eeles,Zsofia Kote-Jarai,Kenneth R Muir,Artitaya Lophatananon,Johanna Schleutker,Nora Pashayan,Jyotsna Batra,Henrik Grönberg,David E Neal,Børge G Nordestgaard,Catherine M Tangen,Robert J MacInnis,Alicja Wolk,Demetrius Albanes,Christopher A Haiman,Ruth C Travis,William J Blot,Janet L Stanford,Lorelei A Mucci,Catharine M L West,Sune F Nielsen,Adam S Kibel,Olivier Cussenot,Sonja I Berndt,Stella Koutros,Karina Dalsgaard Sørensen,Cezary Cybulski,Eli Marie Grindedal,Florence Menegaux,Jong Y Park,Sue A Ingles,Christiane Maier,Robert J Hamilton,Barry S Rosenstein,Yong-Jie Lu,Stephen Watya,Ana Vega,Manolis Kogevinas,Fredrik Wiklund,Kathryn L Penney,Chad D Huff,Manuel R Teixeira,Luc Multigner,Robin J Leach,Hermann Brenner,Esther M John,Radka Kaneva,Christopher J Logothetis,Susan L Neuhausen,Kim De Ruyck,Piet Ost,Azad Razack,Lisa F Newcomb,Jay H Fowke,Marija Gamulin,Aswin Abraham,Frank Claessens,Jose Esteban Castelao,Paul A Townsend,Dana C Crawford,Gyorgy Petrovics,Ron H N van Schaik,Marie-Élise Parent,Jennifer J Hu,Wei Zheng,UKGPCS collaborators,APCB (Australian Prostate Cancer BioResource),NC-LA PCaP Investigators,IMPACT Study Steering Committee and Collaborators,Canary PASS Investigators,Profile Study Steering Committee,PRACTICAL Consortium,Ian G Mills,Ole A Andreassen,Anders M Dale,Tyler M Seibert
DOI: https://doi.org/10.1038/s41391-022-00497-7
Abstract:Background: Prostate cancer risk stratification using single-nucleotide polymorphisms (SNPs) demonstrates considerable promise in men of European, Asian, and African genetic ancestries, but there is still need for increased accuracy. We evaluated whether including additional SNPs in a prostate cancer polygenic hazard score (PHS) would improve associations with clinically significant prostate cancer in multi-ancestry datasets. Methods: In total, 299 SNPs previously associated with prostate cancer were evaluated for inclusion in a new PHS, using a LASSO-regularized Cox proportional hazards model in a training dataset of 72,181 men from the PRACTICAL Consortium. The PHS model was evaluated in four testing datasets: African ancestry, Asian ancestry, and two of European Ancestry-the Cohort of Swedish Men (COSM) and the ProtecT study. Hazard ratios (HRs) were estimated to compare men with high versus low PHS for association with clinically significant, with any, and with fatal prostate cancer. The impact of genetic risk stratification on the positive predictive value (PPV) of PSA testing for clinically significant prostate cancer was also measured. Results: The final model (PHS290) had 290 SNPs with non-zero coefficients. Comparing, for example, the highest and lowest quintiles of PHS290, the hazard ratios (HRs) for clinically significant prostate cancer were 13.73 [95% CI: 12.43-15.16] in ProtecT, 7.07 [6.58-7.60] in African ancestry, 10.31 [9.58-11.11] in Asian ancestry, and 11.18 [10.34-12.09] in COSM. Similar results were seen for association with any and fatal prostate cancer. Without PHS stratification, the PPV of PSA testing for clinically significant prostate cancer in ProtecT was 0.12 (0.11-0.14). For the top 20% and top 5% of PHS290, the PPV of PSA testing was 0.19 (0.15-0.22) and 0.26 (0.19-0.33), respectively. Conclusions: We demonstrate better genetic risk stratification for clinically significant prostate cancer than prior versions of PHS in multi-ancestry datasets. This is promising for implementing precision-medicine approaches to prostate cancer screening decisions in diverse populations.