Polygenic risk score-analysis of thromboembolism in patients with acute lymphoblastic leukemia

Applying polygenic risk score methods to pharmacogenomics GWAS: challenges and opportunities

Polygenic risk scores (PRSs) have emerged as promising tools for the prediction of human diseases and complex traits in disease genome-wide association studies (GWAS). Applying PRSs to pharmacogenomics (PGx) studies has begun to show great potential for improving patient stratification and drug response prediction. However, there are unique challenges that arise when applying PRSs to PGx GWAS beyond those typically encountered in disease GWAS (e.g. Eurocentric or trans-ethnic bias). These challenges include: (i) the lack of knowledge about whether PGx or disease GWAS/variants should be used in the base cohort (BC); (ii) the small sample sizes in PGx GWAS with corresponding low power and (iii) the more complex PRS statistical modeling required for handling both prognostic and predictive effects simultaneously. To gain insights in this landscape about the general trends, challenges and possible solutions, we first conduct a systematic review of both PRS applications and PRS method development in PGx GWAS. To further address the challenges, we propose (i) a novel PRS application strategy by leveraging both PGx and disease GWAS summary statistics in the BC for PRS construction and (ii) a new Bayesian method (PRS-PGx-Bayesx) to reduce Eurocentric or cross-population PRS prediction bias. Extensive simulations are conducted to demonstrate their advantages over existing PRS methods applied in PGx GWAS. Our systematic review and methodology research work not only highlights current gaps and key considerations while applying PRS methods to PGx GWAS, but also provides possible solutions for better PGx PRS applications and future research.

A systematic review and analysis of the use of polygenic scores in pharmacogenomics

Polygenic scores (PGS) have emerged as promising tools for complex trait risk prediction. The application of these scores to pharmacogenomics provides new opportunities to improve the prediction of treatment outcomes. To gain insight into this area of research, we conducted a systematic review and accompanying analysis. This review uncovered 51 papers examining the use of PGS for drug-related outcomes, with the majority of these papers focusing on the treatment of psychiatric disorders (n=30). Due to difficulties in collecting large cohorts of uniformly treated patients, the majority of pharmacogenomic PGS were derived from large-scale genome-wide association studies of disease phenotypes that were related to the pharmacogenomic phenotypes under investigation (e.g. schizophrenia-derived PGS for antipsychotic response prediction). Examination of the research participants included in these studies revealed that the majority of cohort participants were of European descent (78.4%). These biases were also reflected in research affiliations, which were heavily weighted towards institutions located in Europe and North America, with no first or last authors originating from institutions in Africa or South Asia. There was also substantial variability in the methods used to develop PGS, with between 3 and 6.6 million variants included in the PGS. Finally, we observed significant inconsistencies in the reporting of PGS analyses and results, particularly in terms of risk model development and application, coupled with a lack of data transparency and availability, with only three pharmacogenomics PGS deposited on the PGS Catalog. These findings highlight current gaps and key areas for future pharmacogenomic PGS research.

Endothelial dysfunction and thromboembolism in children, adolescents, and young adults with acute lymphoblastic leukemia

Endothelial dysfunction has not previously been investigated as a thrombogenic risk factor among patients with acute lymphoblastic leukemia (ALL), known to be at high risk of thromboembolism. We retrospectively explored the association between three circulating biomarkers of endothelial dysfunction (thrombomodulin, syndecan-1, VEGFR-1) measured in prospectively collected blood samples and risk of thromboembolism in 55 cases and 165 time-matched controls, treated according to the NOPHO ALL2008 protocol. In age-, sex-, and risk group-adjusted analysis, increasing levels of thrombomodulin and VEGFR-1 were independently associated with increased odds of developing thromboembolism (OR 1.37 per 1 ng/mL [95% CI 1.20‒1.56, P < 0.0001] and OR 1.21 per 100 pg/mL [95% CI 1.02‒1.21, P = 0.005], respectively). These associations remained significant when including only samples drawn >30 days before thromboembolic diagnosis. Thrombomodulin levels were on average 3.2 ng/mL (95% CI 2.6-8.2 ng/mL) higher in samples with measurable asparaginase activity (P < 0.0001). Among single nucleotide variants located in or neighboring coding genes for the three biomarkers, none were significantly associated with odds of thromboembolism. If results are validated in another cohort, thrombomodulin and VEGFR-1 could serve as predictive biomarkers, identifying patients in need of preemptive antithrombotic prophylaxis.