Can polygenic risk scores contribute to cost-effective cancer screening? A systematic review

Abstract

Polygenic risk is known to influence susceptibility to cancer. The use of data on polygenic risk, in conjunction with other predictors of future disease status, may offer significant potential for preventative care through risk-stratified screening programmes. An important element in the evaluation of screening programmes is their cost-effectiveness.We undertook a systematic review of papers evaluating the cost-effectiveness of screening interventions informed by polygenic risk scores compared to more conventional screening modalities. We included papers reporting cost-effectiveness outcomes in the English language published as articles or uploaded onto preprint servers with no restriction on date, type of cancer or form of polygenic risk modelled. We excluded papers evaluating screening interventions that did not report cost-effectiveness outcomes or which had a focus on monogenic risk. We evaluated studies using the Quality of Health Economic Studies checklist.Ten studies were included in the review, which investigated three cancers: prostate (n=5), colorectal (n=3) and breast (n=2). All study designs were cost-utility papers implemented as Markov models (n=6) or microsimulations (n=4). Nine of ten papers scored highly (score >75 on a 0-100) scale) when assessed using the Quality of Health Economic Studies checklist. Eight of ten studies concluded that polygenic risk informed cancer screening was likely to be more cost-effective than alternatives. However, the included studies lacked robust external data on the cost of polygenic risk stratification, did not account for how very large volumes of polygenic risk data on individuals would be collected and used, did not consider ancestry-related differences in polygenic risk, and did not fully account for downstream economic sequalae stemming from the use of polygenic risk data in these ways. These topics merit attention in future research on how polygenic risk data might contribute to cost-effective cancer screening.FundingThis work was supported by Cancer Research UK under grant number C18281/A29019. PD and RM are members of the MRC Integrative Epidemiology Unit at the University of Bristol which is supported by the Medical Research Council and the University of Bristol (MC_UU_12013/1, MC_UU_12013/9). PD, EK ad RMM received support from a Cancer Research UK (C18281/A29019) programme grant (the Integrative Cancer Epidemiology Programme). SW and JCT receive funding from the Oxford NIHR Biomedical Research Centre.RegistrationProspectively registered on PROSPERO database before searches commenced. Available at https://www.crd.york.ac.uk/prospero/display_record.php?RecordID=243659