Abstract
It is well accepted that GD (8,9), HT (9,10), and subclinical AITD (9,11,12), along with other common thyroid diseases such as goiter (13,14) and thyroid nodularity (15), are complex diseases. They are all multifactorial, with the clinical phenotype representing the net effect of many contributing genetic and environmental factors. It has been difficult to separate environmental influences from genetic susceptibility. Progress with respect to revealing the exact chromosomal localization of the heritable fraction and the relevant environmental triggers has been slow. Very little is known of the roles of gene–environment and gene–gene interactions and much remains to be learned about how best to quantify this.
A role for genetic factors in the etiology of AITD has long been recognized. Familial clustering was illustrated by the finding of 2–15 times higher prevalence of AITD among first degree relatives of patients with AITD as compared with the background population (16 –19). Subsequent twin studies (8,10,12,20) demonstrated significantly higher concordance rates in monozygotic than in fraternal twins, proving that a significant part of the familial aggregation is due to shared genes. Using structural equation modeling, we have estimated that the heritability of GD (8) and subclinical AITD (12) is 79% (95% CI: 38%–90%) and 73% (95% CI: 46%–89%), respectively.
A detailed account of the identification, localization, and characterization of the genetic contribution to the development of AITD is beyond the scope of this editorial, but these issues have been dealt with in recent publications (21 –23). In brief, AITD is polygenic, with no single gene being either necessary or sufficient for disease development. The genes identified so far can be divided into two major groups: the immune-modulating genes (i.e. HLA-, CTLA4, and PTPN22) and the thyroid-specific genes, reflected by thyroglobulin and thyrotropin receptor. Some genes confer susceptibility to both GD and HT, whereas others are specifically related to the risk of only one phenotype within the AITD complex. Moreover, association data demonstrate a considerable heterogeneity between different ethnic groups (22,23). More perplexing is the finding, although based on limited data, that the genetic background may vary between families (6) and one study has very recently indicated the possibility of a sex-specific genetic locus (24).
The idea of environmental exposures as the triggering event in the development of clinical and subclinical AITD has long been favored. The fact that monozygotic twin pairs show incomplete concordance rates for GD (8) and HT (10) as well as for subclinical AITD (12) provides evidence of environmental factors being of importance in the etiology of these phenotypes. In fact, the incomplete concordance within monozygotic twin pairs highlights that in the absence of critical environmental exposure acting on a particular genotype, disease will not develop. In accordance, biometric twin modeling has shown that 20%–30% of the phenotypic variation in AITD is explained by environmental factors (8,12). In line with this, a considerable number of more or less well-characterized specific environmental factors, such as cigarette smoking, stress, low birth weight, level of iodine intake, selenium deficiency, viral and bacterial infections, and microchimerism, to name but a few, have been associated with AITD (9,25,26). It appears that multiple environmental factors are involved in the development of AITD, with no single factor being either necessary or sufficient for disease development. Even more bewildering is the finding that sometimes a specific environmental factor seems to have opposing effects depending on the type of AITD. For instance, cigarette smoking is a risk factor for the development of GD (27) especially Graves' ophthalmopathy (28). In contrast, smoking has been associated with a lower risk of HT and subclinical AITD (29,30).
The pronounced sex difference in prevalence of AITD, often 5–10 times more common in females, is puzzling. Theoretically, a skewed X chromosome inactivation (XCI) pattern resulting in tissue chimerism offers a biologically plausible explanation for this phenomenon. The X chromosome harbors several genes crucial for maintenance of immune function and tolerance (31) as well as genetic markers in linkage or association with AITD (32 –34). Therefore, a skewed XCI (i.e., predominant inactivation of the normal X chromosome) could result in expression of X-linked disorders. We (35), and subsequently others (36,37), have reported an association between a skewed XCI pattern and the presence of clinically overt AITD in females. With regard to subclinical AITD and the XCI pattern, the association was attenuated and nonsignificant within monozygotic twin pairs, indicating that XCI and subclinical AITD are, at least partly, influenced by common genetic determinants (i.e., genetic confounding) (38). Recently, fetal microchimerism, which reflects the presence of small populations of cells from one individual (fetus) in another genetically distinct individual (mother), has also been suggested as a possible explanation for the observed female preponderance of AITD. This hypothesis is strengthened by the finding of microchimeric cell populations within the thyroid gland of women with HT and GD (39,40). In support of this, a higher prevalence of thyroid autoantibodies have been reported in both female and male twins from fraternal pairs, as opposed to monozygotic pairs, indicating a potential role of microchimerism in subclinical AITD as well (41). At variance with this are two large studies failing to demonstrate an association between parity and overt and/or subclinical AITD, suggesting a limited role, if any, for microchimerism in the etiology of AITD (42,43). Intake of oral contraceptives, another potential explanation for the sex-related difference in the frequency of AITD, is actually linked with a decreased prevalence of TPOAb (44) as well as a lower prevalence of overt thyroid dysfunction (45).
The challenge faced by research into the etiology of complex phenotypes, such as those for AITD, is to identify genes and environmental factors having a relatively small effect against a background of substantial genetic and environmental diversity. Additionally, it has become evident that each patient with AITD harbors a unique cluster of genetic and environmental susceptibility factors, and only very few are shared with other individuals with the same clinical label (9,23). It follows that the true effect of both heredity and environmental factors in the development of AITD can only be discerned when examined in conjunction. Unfortunately, there is an overwhelming deficit of adequately powered longitudinal studies in genetically well-characterized individuals in whom to determine how and when well-defined environmental exposures—such as level of iodine intake, Yersinia enterocolitica infection or other infections, or the quantity of cigarettes smoked—influence the development of AITD. Moreover, with very few exceptions, the impact of various genetic markers as well as environmental triggers on disease development has only been investigated in isolation and without acknowledging the fact that both the genetic setup and environmental exposures change over time.
In the present issue of Thyroid, Hou et al. (46) evaluate in a prospective manner the development of thyroid dysfunction and thyroid autoantibodies as well as the impact of cigarette smoking on the development of these events in a cohort of first degree relatives of Chinese Han patients with GD. In line with what has been reported in small white cohorts, this study confirms the presence of genetic anticipation in GD (47,48), as reflected by a generational decrease in age of disease onset. In this Asian population Hou et al. (46) also convincingly demonstrate familial aggregation of thyroid autoantibodies (11,49) and provide some evidence in favor of a protective effect of smoking on the development of subclinical AITD (30).
Although the cohort used in the present study by Hou et al. (46) has been genetically characterized (50) the authors unfortunately, and surprisingly, do not utilize this information. Had they done so, the unique genetic information could have provided valuable insight into how modifiable environmental triggers, such as iodine intake or smoking (or both in combination), interact with the suspected genetic markers in influencing final phenotype. At present, the known environmental risk factors in AITD, as in most chronic multifactorial diseases, have a very poor predictive value for disease development. Therefore, stratifying according to genetic susceptibility, with respect to one or more loci, could potentially have significantly improved the predictive value for disease development among biologically susceptible individuals. Accumulating such knowledge would eventually, at least on a population level, increase our knowledge of the etiology of AITD. Given that AITD occurs based on a unique combination of genetic, epigenetic, and environmental factors, it can be questioned whether this information will be sensitive and specific enough to allow cost-effective prevention and therapy in the individual patient.