The genetics of Graves' disease

Abstract

Graves' disease (GD) is the commonest cause of hyperthyroidism and has a strong female preponderance. Everyday clinical practice suggests strong aggregation within families and twin studies demonstrate that genetic factors account for 60-80% of risk of developing GD. In this review, we collate numerous genetic studies and outline the discoveries over the years, starting with historic candidate gene studies and then exploring more recent genome-wide linkage and association studies, which have involved substantial cohorts of East Asian patients as well as those of European descent. Variants in genes including HLA, CTLA4, and PTPN22 have been shown to have substantial individual effects on disease susceptibility. In addition, we examine emerging evidence concerning the possibility that genetic variants may correlate with relevant clinical phenotypes including age of onset of GD, severity of thyrotoxicosis, goitre size and relapse of hyperthyroidism following antithyroid drug therapy, as well as thyroid eye disease. This review supports the inheritance of GD as a complex genetic trait, with a growing number of more than 80 susceptibility loci identified so far. Future implementation of more targeted clinical therapies requires larger studies investigating the influence of these genetic variants on the various phenotypes and different outcomes of conventional treatments.

Keywords: Autoimmune thyroid disease; Genome-wide association; Genotype-phenotype correlation; Graves’ disease; Immunogenetics; Thyroid eye disease.

Fig. 1

Genome-wide association scan in 30,000 patients with autoimmune thyroid disease- top 10 susceptibility loci. Data taken from Saevarsdottir et al. [58], show the balance between the odds ratio of the susceptibility allele predisposing to disease vs the allele frequency in the population. Bubble size/colour represents the probability in favour of association (-log10 P value). It is notable that probability for association is less for alleles with very low (FLT3, ADCY7) and very high (IFIH1, TYK2) population frequency, owing to low power.

Fig. 2

The GREAT+ score divides patients with Graves’ disease according to age, presenting serum free thyroxine concentration, TRAb antibody titre, and genotype for HLA class II variants, CTLA4 ‘CT60’ (rs3087243) and PTPN22 R620W (rs2476601) alleles (from Vos et al. reference 64)