Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2024 Oct 3;26(10):1933-1944.
doi: 10.1093/neuonc/noae112.

Genome-wide polygenic risk scores predict risk of glioma and molecular subtypes

Taishi Nakase  1 , Geno A Guerra  2 , Quinn T Ostrom  3 , Tian Ge  4   5   6 , Beatrice S Melin  7 , Margaret Wrensch  2 , John K Wiencke  2 , Robert B Jenkins  8 , Jeanette E Eckel-Passow  9 ; Glioma International Case-Control Study (GICC); Melissa L Bondy  1   10 , Stephen S Francis  2   11   12 , Linda Kachuri  1   10
Affiliations

Genome-wide polygenic risk scores predict risk of glioma and molecular subtypes

Taishi Nakase et al. Neuro Oncol. .

Abstract

Background: Polygenic risk scores (PRS) aggregate the contribution of many risk variants to provide a personalized genetic susceptibility profile. Since sample sizes of glioma genome-wide association studies (GWAS) remain modest, there is a need to efficiently capture genetic risk using available data.

Methods: We applied a method based on continuous shrinkage priors (PRS-CS) to model the joint effects of over 1 million common variants on disease risk and compared this to an approach (PRS-CT) that only selects a limited set of independent variants that reach genome-wide significance (P < 5 × 10-8). PRS models were trained using GWAS stratified by histological (10 346 cases and 14 687 controls) and molecular subtype (2632 cases and 2445 controls), and validated in 2 independent cohorts.

Results: PRS-CS was generally more predictive than PRS-CT with a median increase in explained variance (R2) of 24% (interquartile range = 11-30%) across glioma subtypes. Improvements were pronounced for glioblastoma (GBM), with PRS-CS yielding larger odds ratios (OR) per standard deviation (SD) (OR = 1.93, P = 2.0 × 10-54 vs. OR = 1.83, P = 9.4 × 10-50) and higher explained variance (R2 = 2.82% vs. R2 = 2.56%). Individuals in the 80th percentile of the PRS-CS distribution had a significantly higher risk of GBM (0.107%) at age 60 compared to those with average PRS (0.046%, P = 2.4 × 10-12). Lifetime absolute risk reached 1.18% for glioma and 0.76% for IDH wildtype tumors for individuals in the 95th PRS percentile. PRS-CS augmented the classification of IDH mutation status in cases when added to demographic factors (AUC = 0.839 vs. AUC = 0.895, PΔAUC = 6.8 × 10-9).

Conclusions: Genome-wide PRS has the potential to enhance the detection of high-risk individuals and help distinguish between prognostic glioma subtypes.

Keywords: genetic susceptibility; glioma; polygenic risk score (PRS); prediction; risk.

PubMed Disclaimer

Conflict of interest statement

The authors declare no competing interests.

Figures

Figure 1.
Figure 1.
Overview of the study design.
Figure 2.
Figure 2.
Prediction accuracy of PRS-CT and PRS-CS in the UK Biobank and TCGA/WTCCC datasets across glioma subtypes. (A and B) Prediction accuracy, measured as variance explained (R2) on the liability scale, for PRS trained using discovery GWAS summary statistics specific to target phenotype (histological: all glioma, GBM, non-GBM; molecular: IDH wildtype, IDH mutant). Each bar represents 1 testing cohort with the shade of the bar representing the testing dataset (light = UK Biobank; dark = TCGA/WTCCC). (C and D) The OR per SD unit increase in PRS for each glioma subtype in the UK Biobank and TCGA/WTCCC testing datasets. Each dot corresponds to 1 testing cohort with the shape of the dot representing the PRS construction method (circle = PRS-CT; triangle = PRS-CS) and the shade of the dot representing the testing dataset (light = UK Biobank; dark = TCGA/WTCCC). The error bars represent the 95% CI. The R2, OR, and 95% CI for each subtype, PRS construction method and testing cohort are provided in Supplementary Tables 5–7.
Figure 3.
Figure 3.
Relative glioma risk by PRS category stratified by histological subtype. (A) All glioma. (B) GBM subtype. (C) Non-GBM subtype. The x-axis indicates the percentiles of the PRS distribution (0–40%, 40–60%, 60–80%, 80–100%, and 95–100%). The y-axis indicates odds ratios (ORs) with error bars representing 95% CI for each PRS category relative to the middle category (40–60%) of risk scores. The results are stratified by testing dataset (light = UK Biobank; dark = TCGA/WTCCC) and PRS method (circle = PRS-CT; triangle = PRS-CS). Each PRS (colored shape) is trained using discovery GWAS summary statistics that correspond to the target phenotype.
Figure 4.
Figure 4.
Comparison of PRS performance for logistic regression models that classify TCGA cases according to IDH mutation status. Classification accuracy of TCGA cases (384 IDH mutant; 384 IDH wildtype) for PRS-CT and PRS-CS using PRS trained on discovery GWAS summary statistics with histological (GBM or non-GBM) or molecular profiling (IDH mutant and IDH wildtype). Each bar represents the area under the receiver operating characteristic curve (AUC) and 95% CI for a single logistic regression model adjusted for the first 10 genetic ancestry principal components with the shade of the bar representing the PRS construction method (light = PRS-CT; dark = PRS-CS). The x-axis indicates the PRS included in each model (single: GBM, non-GBM, IDH wildtype, IDH mutant; multiple: GBM + non-GBM, IDH mutant + IDH wildtype, all = GBM + non-GBM + IDH mutant + IDH wildtype). Results with additional covariates are provided in Supplementary Tables 9 and 10.
Figure 5.
Figure 5.
Estimated 5-year cumulative incidence as a function of age stratified by percentiles of the PRS-CS distribution for a typical individual of European ancestry in the UK Biobank. Estimates and corresponding 95% CI were obtained from cause-specific Cox regression models using incident glioma cases for all glioma (N = 659), GBM (N = 505) and non-GBM (N = 93), and cancer-free controls (N = 412 556). Low PRS corresponds to below the 20th percentile, average PRS is defined as between the 20th and 80th percentiles and high PRS includes individuals above the 80th percentile of the normalized PRS-CS distribution. Each PRS-CS is trained on the discovery GWAS summary statistics corresponding to the target phenotype (all glioma, GBM, and non-GBM). The shaded areas represent 95% CI.
Figure 6.
Figure 6.
Lifetime absolute risks for each glioma subtype by PRS category. (A) Histological subtypes (all glioma, GBM, and non-GBM). (B) Molecular subtypes (IDH wildtype and IDH mutant). The x-axis indicates the PRS percentile, and the y-axis is the lifetime absolute risk with error bars for the 95% CI. Each PRS (colored dots) is trained on discovery GWAS summary statistics that correspond to the target phenotype (labeled grey bar) using PRS-CS with the shading of the dots representing the testing cohort (light = UK Biobank; dark = TCGA/WTCCC).
See this image and copyright information in PMC

Update of

Similar articles

See all similar articles

Cited by

References

    1. Ostrom QT, Price M, Neff C, et al. CBTRUS statistical report: primary brain and other central nervous system tumors diagnosed in the United States in 2016—2020. Neuro-Oncology. 2023;25(12 Suppl 2):iv1–iv99. - PMC - PubMed
    1. Molinaro AM, Taylor JW, Wiencke JK, Wrensch MR.. Genetic and molecular epidemiology of adult diffuse glioma. Nat Rev Neurol. 2019;15(7):405–417. - PMC - PubMed
    1. Louis DN, Perry A, Wesseling P, et al. The 2021 WHO classification of tumors of the central nervous system: a summary. Neuro-Oncology. 2021;23(8):1231–1251. - PMC - PubMed
    1. Poon MTC, Sudlow CLM, Figueroa JD, Brennan PM.. Longer-term (≥2 years) survival in patients with glioblastoma in population-based studies pre- and post-2005: a systematic review and meta-analysis. Sci Rep. 2020;10(1):11622. - PMC - PubMed
    1. Scheurer ME, Etzel CJ, Liu M, et al. Aggregation of cancer in first-degree relatives of patients with glioma. Cancer Epidemiol Biomarkers Prev 2007;16(11):2491–2495. - PMC - PubMed

Substances