Integration of genetic risk factors into a clinical algorithm for multiple sclerosis susceptibility: a weighted genetic risk score

Abstract

Background: Prediction of susceptibility to multiple sclerosis (MS) might have important clinical applications, either as part of a diagnostic algorithm or as a means to identify high-risk individuals for prospective studies. We investigated the usefulness of an aggregate measure of risk of MS that is based on genetic susceptibility loci. We also assessed the added effect of environmental risk factors that are associated with susceptibility for MS.

Methods: We created a weighted genetic risk score (wGRS) that includes 16 MS susceptibility loci. We tested our model with data from 2215 individuals with MS and 2189 controls (derivation samples), a validation set of 1340 individuals with MS and 1109 controls taken from several MS therapeutic trials (TT cohort), and a second validation set of 143 individuals with MS and 281 controls from the US Nurses' Health Studies I and II (NHS/NHS II), for whom we also have data on smoking and immune response to Epstein-Barr virus (EBV).

Findings: Individuals with a wGRS that was more than 1.25 SD from the mean had a significantly higher odds of MS in all datasets. In the derivation sample, the mean (SD) wGRS was 3.5 (0.7) for individuals with MS and 3.0 (0.6) for controls (p<0.0001); in the TT validation sample, the mean wGRS was 3.4 (0.7) for individuals with MS versus 3.1 (0.7) for controls (p<0.0001); and in the NHS/NHS II dataset, the mean wGRS was 3.4 (0.8) for individuals with MS versus 3.0 (0.7) for controls (p<0.0001). In the derivation cohort, the area under the receiver operating characteristic curve (C statistic; a measure of the ability of a model to discriminate between individuals with MS and controls) for the genetic-only model was 0.70 and for the genetics plus sex model was 0.74 (p<0.0001). In the TT and NHS cohorts, the C statistics for the genetic-only model were both 0.64; adding sex to the TT model increased the C statistic to 0.72 (p<0.0001), whereas adding smoking and immune response to EBV to the NHS model increased the C statistic to 0.68 (p=0.02). However, the wGRS does not seem to be correlated with the conversion of clinically isolated syndrome to MS.

Interpretation: The inclusion of 16 susceptibility alleles into a wGRS can modestly predict MS risk, shows consistent discriminatory ability in independent samples, and is enhanced by the inclusion of non-genetic risk factors into the algorithm. Future iterations of the wGRS might therefore make a contribution to algorithms that can predict a diagnosis of MS in a clinical or research setting.

Figure 1. Distribution of wGRS in cases and controls

A wGRS is calculated for each subject within each of the studied cohorts. Here, we illustrate the distribution of wGRS in the derivation samples, with cases in dark blue and controls in light blue. The TT sample collection is overlaid with cases in dark green and controls in light green. Finally, the Czech SET collection of CIS cases is plotted in red.

Figure 2. Odds ratios for risk categories defined using the wGRS. (a) Derivation samples

Data is presented only for subjects with MS. Seven categories of genetic risk are defined using the control subjects in the derivation samples, with “1” being the lowest risk category. The distribution of MS subjects amongst the seven risk categories is plotted in black as a histogram and is skewed since these subjects have a greater risk of MS than do healthy subjects. In red, we superimpose the log of the odds ratio (red triangle) for MS susceptibility of each risk category, along with a 95% confidence interval for that estimate. (b and c) TT and NHS/NHS II validation samples. Here, we present the distribution of wGRS in the seven categories of risk defined in each of the two sample collections used in validation exercises.

Figure 2. Odds ratios for risk categories defined using the wGRS. (a) Derivation samples

Data is presented only for subjects with MS. Seven categories of genetic risk are defined using the control subjects in the derivation samples, with “1” being the lowest risk category. The distribution of MS subjects amongst the seven risk categories is plotted in black as a histogram and is skewed since these subjects have a greater risk of MS than do healthy subjects. In red, we superimpose the log of the odds ratio (red triangle) for MS susceptibility of each risk category, along with a 95% confidence interval for that estimate. (b and c) TT and NHS/NHS II validation samples. Here, we present the distribution of wGRS in the seven categories of risk defined in each of the two sample collections used in validation exercises.

Figure 3. ROC curves for models predicting a diagnosis of MS or CIS. (a) Derivation samples

We plot the results for three separate models predicting a diagnosis of MS: “GRS w/o HLA DRB1” that includes 15 susceptibility loci and excludes the HLA DRB1*1501 allele (blue), “GRS w/HLA DRB1” that includes 15 susceptibility loci and the HLA DRB1*1501 allele (red), “GRS w/HLA DRB1+ female” that includes 15 susceptibility loci, the HLA DRB1*1501 allele, and gender (green). (b) TT samples. We repeat the analysis in (a) using the TT validation samples. The same presentation scheme is used. (c) NHS/NHS II samples. The analysis method is the same in the NHS/NHS II validation samples. These subjects are all women. The third model that is plotted (green line) includes all 15 susceptibility loci, the HLA DRB1*1501 allele, as well as terms for smoking and EBV titers.

Figure 3. ROC curves for models predicting a diagnosis of MS or CIS. (a) Derivation samples

We plot the results for three separate models predicting a diagnosis of MS: “GRS w/o HLA DRB1” that includes 15 susceptibility loci and excludes the HLA DRB1*1501 allele (blue), “GRS w/HLA DRB1” that includes 15 susceptibility loci and the HLA DRB1*1501 allele (red), “GRS w/HLA DRB1+ female” that includes 15 susceptibility loci, the HLA DRB1*1501 allele, and gender (green). (b) TT samples. We repeat the analysis in (a) using the TT validation samples. The same presentation scheme is used. (c) NHS/NHS II samples. The analysis method is the same in the NHS/NHS II validation samples. These subjects are all women. The third model that is plotted (green line) includes all 15 susceptibility loci, the HLA DRB1*1501 allele, as well as terms for smoking and EBV titers.