Organ-specific autoimmune disorders

Several large recent studies have clarified the strength of the association between AITD and type 1 diabetes mellitus. The prevalence of thyroid peroxidase and/or TG antibodies in 197 recent onset German diabetic patients was 18.4%, 7.8% in their first degree relatives and 3.2% in 150 controls, indicating a strong genetic component to this association (4). Generally higher figures of 27% (n = 109), 28% and 10% respectively were reported in similar patient groups from Turkey (5). Up to half of the diabetic children with thyroid antibodies in two studies from Germany and Denmark developed elevated TSH levels or goitre (6), or an abnormal echogenic pattern on thyroid ultrasound (7), within 4 years of follow-up. Diabetic patients who are older, female, have concurrent parietal cell antibodies and are glutamic acid decarboxylase antibody-positive have the greatest risk of AITD (8). Latent autoimmune diabetes of adults (LADA) is also associated with a 2-fold increase in TPO antibody prevalence, compared to matched type 2 diabetic patients (9).

The strong association of AITD with autoimmune Addison’s disease has been long documented (10). Details of a uniquely large Italian cohort have recently been published; of over 4000 patients with AITD , 1% had antibodies to the adrenal cortex (11). In a series of 94 nationally recruited Norwegian Addison’s patients, 29% had autoimmune hypothyroidism and 6% had Graves’ disease (12).

Tissue transglutaminase antibodies were found in 3.2% of 220 Italian patients with AITD, and all had coeliac disease on biopsy; even given recent reports of a population prevalence of 1%, this represents at least a 3-fold excess of coeliac disease (13). A huge survey of 2624 vitiligo patients from N.America and the UK found that 17% had AITD (21% in women, 6% in men), compared with 5.7% of first degree relatives and an expected control population prevalence of 1.9% (14). These figures are similar to those above for diabetic families. In a much smaller series of 106 Austrian vitiligo patients, AITD was found in 21% compared to 3% of controls; of interest, the presence of antinuclear antibodies in these patients was associated with atrophic thyroiditis (15). In 200 UK patients with primary biliary cirrhosis, 21% had clinically significant AITD and 34% had thyroid autoantibodies (16). Taking into account that the famous Wickham study of the prevalence of AITD in the healthy population originated from the same area, and found much lower figures of AITD in healthy subjects (17), it seems so far that this is a strong and validated association with primary biliary cirrhosis.

The relationship between AITD and multiple sclerosis is much less clear, yet more intriguing in the light of the frequent emergence of Graves’ disease after treatment of the disorder with lymphocyte-depleting monoclonal antibodies (18). In a survey of 571 UK patients with multiple sclerosis, an excess of associated autoimmune disorders was found in the first-degree relatives, especially those with multiplex families (27%, compared to 15.7% in simplex families and 11.7% in controls); AITD made the major contribution to these associated disorders but was not increased in the probands themselves (19). Furthermore, there was no excess of subclinical AITD or TPO antibodies in 152 Italian patients with relapsing-remitting multiple sclerosis, compared to 437 healthy controls, although there was a trend to an increased frequency of TPO antibodies in the men (20). This trend was confirmed in a larger series of 96 consecutive German men with multiple sclerosis, in whom 9.4% had AITD compared to 1.9% of controls, but there was no increase in AITD in 257 female patients (21).

Lymphocytic hypophysitis (LH) is a rare condition often associated with other autoimmune disorders including AITD (22). The autoantigens involved and the pathogenesis remain obscure, although antibodies to the ubiquitous enzyme, ?-enolase, have been proposed as a marker. These antibodies were detected in 70% of biopsy-proven LH patients, 15% of patients with AITD and 10% of controls (23). Rather similar findings have been reported from Japan, with 22% of Graves’ disease patients, 19% of Hashimoto’s patients and 6% of controls being positive for antipituitary antibodies (24). Unsuspected growth hormone deficiency was found in patients with AITD who had pituitary antibodies (25), suggesting a potentially important and novel association that requires confirmation and further follow-up.

Non-organ-specific autoimmunity

Although less strongly associated than the organ-specific disorders, recent studies have confirmed the increased frequency of AITD in systemic lupus erythematosus (SLE) and rheumatoid arthritis (RA). In 300 SLE patients in the UK, the prevalence of autoimmune hypothyroidism was 5.7% and 14% had thyroid autoantibodies; the prevalence of hyperthyroidism was not increased (26). The frequency of AITD and TPO antibodies in SLE is around twice that seen in RA but in the latter still exceeds that expected in the population (27). Indeed, the association of RA with AITD has led to spurious associations being reported between RA and polymorphisms of the CTLA-4 gene; the association disappears once those RA patients with AITD are removed from analysis (28). Primary Sjögren’s syndrome is also associated with AITD; 30% of 137 patients had AITD compared to 4% of controls in a recent French survey (29) and of 25 Brazilian patients, 52% had thyroid antibodies compared to 4% of controls (30).

The reality of an association between chronic urticaria and AITD has been debated but two recent studies support this association. Of 187 children with chronic urticaria, 4.3% had thyroid antibodies, around three times higher than controls derived from the literature (31). In 45 adults with chronic urticaria, 27% had thyroid antibodies compared to 3% of controls, although none had clinical evidence of thyroid disease (32).

A novel association with AITD was recently reported in 16 patients with chronic periaortitis, in whom 3 (19%) had TPO antibodies, and one of whom had hypothyroidism (33). Replication is clearly needed.

Summary
There are an ever-increasing number of studies confirming expected associations (Table) and demonstrating possible new associations between AITD and other autoimmune diseases.

Summary of main disease associations with AITD

The associations are important clinically, although there is as yet no cost-benefit analysis of screening programmes for these disorders based on the known associations. This is an area ripe for large, multicentre studies as the focus increases on using autoantibodies as predictors of disease and new developments in array technology offer new dimensions in assays for screening (34).