Reference | Year | Study design | Cohort | Cohort location | Sample size | No. of SNPs | AUC (95% CI) | Risk prediction |
---|---|---|---|---|---|---|---|---|
Hüsing et al. [28] | 2012 | Retrospective | Non-familial | USA, Europe | 6009 cases 7827 controls | 32 | 0.58 (0.57–0 .60) | – |
Sawyer et al. [7] | 2012 | Retrospective | High-risk families | Australia | 1143 cases 892 controls | 22 | 0.65 (0.63–0.68) | 4.5-fold increased risk between lowest to highest quartile Lifetime risk: 6 to 27% for lowest and top quartile |
Mavaddat et al. [29] | 2015 | Retrospective | Non-familial | USA, Canada, Australia, Europe | 33,673 cases 33,381 controls | 77 | 0.62 (0.62–0.63) | Threefold increased risk between fist centile and middle quintile Lifetime risk: 3.5 to 29% for lowest centile to highest centile |
Shieh et al. [30] | 2016 | Retrospective | Non-familial | USA Caucasian | 387 cases 387 controls | 83 | 0.59 (0.56–0.53) | Twofold increased risk between lowest and higher quartile |
Asian American | 51 cases 51 controls | 76 | 0.64 (0.53–0.74) | |||||
Evans et al. [31] | 2016 | Retrospective | High-risk families | England | 364 cases 1605 controls | 18 | 0.59 (0.55–0.63) | Twofold increased risk between lowest to highest quintile |
Muranen et al. [32] | 2016 | Retrospective | High-risk families | Finland | 427 cases 1272 controls | 75 | Not reported | 1.55 odds ratio per unit increase in the PRS within the family dataset |
Wen et al. [33] | 2016 | Retrospective | Non-familial | East Asia | 11,760 cases 11,612 controls | 88 | 0.60 (not reported) | 2.26-fold increased risk between lowest and top quartile |
Li et al. [34] | 2017 | Prospective | High-risk families | USA, Australia, Canada | 2869 unaffected 1496 affected | 24 | 0.59 (0.55–0.63) | Threefold increased risk lowest and highest quintile Lifetime risk: 21 to 51% for lowest and highest quintile |
Hsieh et al. [35] | 2017 | Retrospective | Non-familial | Taiwan | 446 cases 514 controls | 6 | 0.60 (not reported) | 2.26-fold increased risk between lowest and highest quartile |
Khera et al. [9] | 2019 | Retrospective | Non-familial | UK Biobank | 6586 cases 157,895 controls | 5218 (LDpred) | 0.69 (0.68–0.69) | – |
Chan et al. [36] | 2018 | Retrospective | Non-familial | Singapore-Chinese | 301 cases 243 controls | 51 | 0.57 (0.51–0.61) | 1.88-fold increased risk between the lowest and higher quartile |
Wang et al. [37] | 2018 | Retrospective | Non-familial | African | 1657 cases 2029 controls | 34 | 0.53 (0.51–0.55) | 4.74-fold increased risk between lowest percentile and top percentile |
Mavaddat et al. [38] | 2019 | Prospective | Non-familial | USA Caucasian | 11,428 cases 18,323 controls | 305 | 0.63Â (0.63-0.65) | 4.37-fold increased risk between middle quintile and highest 1% Lifetime risk of ER-positive disease: 2 to 31% for lowest centile to highest centile Lifetime risk of ER-negative disease: 0.55 to 4% for lowest centile to highest centile |