Patients and methods

Patients with thyroid diseases (n = 397, age 30 (13) years; M/F 75/322) were recruited from two referral centres after informed consent. Ninety four age and sex matched healthy subjects (age 30 (13); M/F 24/70) were recruited from hospital visitors only if their thyroid hormones were within the normal range (triiodothyronine 0.8–3.16 nmol/l; thyroxine 64.5–152.0 nmol/l; thyroid stimulating hormone 0.47–5.0 mU/l). None of the subjects was pregnant or had any concurrent illness that could interfere with thyroid function and all were below the age of 60 years. A uniform physical examination was done for all thyroid patients after taking a thorough clinical history. On the basis of clinical findings and laboratory investigations they were classified into three clinical subgroups on consultation with a board of consultants before the estimation of antithyroid antibodies. The clinical subgroups were: AITD, NAITD, and indeterminate (this group included those in whom no decision could be reached about autoimmunity). Thus, AITD comprised Graves' disease, Hashimoto's thyroiditis, atrophic thyroiditis, silent thyroiditis and postpartum thyroiditis; the indeterminate group those in whom possibility of these varieties of autoimmunity could not be clearly excluded on clinical grounds; and the rest of the thyroid diseases were classified as NAITD. After the measurement of antibodies was completed, each patient's profile was reviewed by the same board of consultants and final assignment to the AITD or NAITD group and their mismatch with initial clinical suspicion was verified. The study was carried out after approval by the Institutional Review Board and Ethical Committee of BIRDEM.

For measurement of antimicrosomal and antithyroglobulin antibodies, serum from each subject was preserved at −70°C until assay. Antibodies were measured by enzyme linked immunosorbent assay (ELISA) using a commercial kit (INOVA Diagnostics Inc, San Diego, USA) with the help of a computerised program of a Kc3 biograph. Antibody titre ⩾130 units for antimicrosomal and ⩾150 units for antithyroglobulin antibody were considered positive whereas values >250 units were considered as strongly positive for both the antibodies. Interassay coefficient variance for antimicrosomal and antithyroglobulin were 7%–17% and 8%–15% respectively. Coefficient variance for triiodothyronine, thyroxine, and thyroid stimulating hormone measurement for control subjects were 3.7%, 6.3%, and 0.8% respectively. Subjects positive for any of the two antibodies or for both were considered positive for antithyroid antibodies.

Statistical analysis of data was done by Statistical Program for Social Studies (SPSS for Windows, release 6.0). Data are expressed in frequencies or mean (SEM) unless stated otherwise. Comparisons were done by the χ² test, McNemar's test, Kruskal-Wallis one way analysis of variance, and Pearson's linear correlation as applicable. Only a p value ⩽0.05 was considered significant.

Results

CLINICAL SUSPICION FOR AUTOIMMUNITY AND FREQUENCY OF POSITIVE ANTIBODIES

Among thyroid patients, 19.4% (77/397) were strongly positive, 7.3% (29/397) were weakly positive, and 73.3% (291/397) were negative for antimicrosomal antibody in comparison with only 4.3% (4/94), 2.1% (2/94), and 93.6% (88/94) respectively in the control group (χ² = 17.852; p = 0.000). For the antithyroglobulin antibody, the frequency for strongly positive, weakly positive, and negative was 27.2% (108/397), 6.8% (27/397), and 66.0% (262/397) respectively among thyroid patients in comparison with 8.5% (8/94), 4.3% (4/94), and 87.2% (82/94) in the control group (χ²= 16.616; p = 0.000). It was found that 13.6% of suspected NAITD and 19% of the indeterminate group were strongly positive, and 4.1% of NAITD and 14.3% of the indeterminate group were weakly positive for antimicrosomal antibody. For antithyroglobulin antibody 18.2% were strongly and 5.0% were weakly positive for suspected NAITD and in the indeterminate group 28.1% and 14.3% respectively. However, a good number of those with clinically suspected AITD were negative for antimicrosomal (61.5%) and antithyroglobulin (51.6%) antibodies (table 1).

FREQUENCY OF POSITIVE ANTIBODIES AMONG VARIOUS CLINICAL SUBGROUPS OF THYROID DISORDERS

Table 2 shows the frequencies of positive antibodies among patients of various clinically suspected thyroid diseases. Of the suspected NAITDs, 100% of those with multinodular hypothyroidism, 46.7% with subacute thyroiditis, 40.0% with suspected iodine deficiency goitre, 31.3% with toxic multinodular goitre, 30.8% with non-toxic solitary nodule, 18.5% with simple multinodular goitre, and 19.4% with simple diffuse goitre were positive for antithyroid antibodies. Among the clinically suspected AITDs, 63.0% with Hashimoto's thyroiditis, 36.4% with Graves' disease, and 44.7% with atrophic hypothyroidism were positive for antithyroid antibodies. In the control group, only 12.8% were positive for the antibodies. None of the subjects was positive for antimicrosomal antibody without being positive for antithyroglobulin antibody in the clinically suspected subgroups of thyroid disorders and in the control group. However, in some of the subgroups, antithyroglobulin antibody was positive without being positive for antimicrosomal antibody. Only 3.6% (4/112) were weakly positive for both the antibodies. However, 24.1% (27/112) were weakly positive for antimicrosomal but strongly positive for antithyroglobulin antibody whereas among the subjects who were positive for only antithyroglobulin antibody, most of them (77%; 27/35) were weakly positive for the antibody. Thus there is an overall greater importance of antimicrosomal antibody.

DIAGNOSTIC IMPORTANCE OF ANTIBODIES

As shown in table 3, it was observed that overall 9.1% (36/397) patients were correctly diagnosed by the measurement of antithyroid antibodies. Thirty three out of 36 patients who were incorrectly diagnosed fell into the subgroup of Hashimoto's thyroiditis. Of the clinically suspected subgroups, 100% of those with multinodular goitre, 30.0% with suspected iodine deficiency goitre, 21.4% with simple multinodular goitre, 30.8% with non-toxic solitary nodule, and 12.9% of simple diffuse goitre were found to be mismatched with their final diagnosis. However, in light of the final diagnosis (table 4), prevalence of thyroid autoimmunity among the thyroid patients was 48.36%. Overall sensitivity and specificity of antibody test were 56.25% and 86.83% respectively among patients. On the other hand, sensitivity of antimicrosomal antibody was 47.67% and specificity 93.14%, which was 56.25% and 86.83% respectively for antithyroglobulin antibody.

CORRELATION

Figure 1 shows that the correlation between the two antibodies was highly significant in both thyroid patients (r = 0.977; p = 0.000) and the control group (r = 0.986; p = 0.000).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Relationship between antimicrosomal and antithyroglobulin antibodies (○ = thyroid patients; • = controls; WHO = World Health Organisation).

Discussion

Antithyroid antibodies are good markers for the assessment of thyroid autoimmunity. In an attempt to find the frequency of positive antibodies among thyroid diseases, we have observed that prevalence of autoimmunity in the patients with thyroid diseases is quite high in Bangladesh. Moreover, autoimmunity in an appreciable number of thyroid patients may remain undetected unless they are checked for antithyroid antibodies.

Compared with healthy controls, thyroid patients were found to be more frequently positive for antithyroid antibodies. Though antithyroid antibodies were more frequently positive among Hashimoto's thyroiditis, atrophic hypothyroidism and Graves' disease, which are known AITDs, quite a few patients in these subgroups were found to be negative. On the other hand, all the multinodular hypothyroid patients as well as many of those with toxic multinodular goitre, subacute thyroiditis, suspected iodine deficiency goitre, and non-toxic solitary nodule were also positive for antithyroid antibodies. It was interesting to observe that many patients with suspected NAITDs and in the indeterminate group were positive for antibodies. In light of the above facts, it is clear that being negative for antibodies does not necessarily exclude thyroid autoimmunity in many instances; but when antibodies are positive, it strongly indicates the autoimmune nature of the disease. Immunological interference has also been found by some studies abroad in NAITD.10-12 21 22 The autoimmune phenomenon should not differ between healthy controls and those with NAITD living under the same environmental conditions; again this would indicate that these antibody positive patients were suffering from AITDs. It is worth mentioning that we have observed positive antithyroid antibodies more frequently among patients with multinodular and diffuse goitre (43% and 36% respectively) as well as among hyperthyroid (37%) and hypothyroid (55%) patients. On the other hand, both mean and median urinary iodine were found higher in multinodular and diffuse goitre as well as in hyperthyroid patients; also both of these were higher among the thyroid patients than controls. However, in none of the groups were these below the cut off value (10 μg/dl) for iodine deficiency (unpublished data).

Antithyroid antibodies are good markers of thyroid autoimmunity.1-4 The relationship between the two antibodies was highly significant; this was true for both the thyroid patients and control subjects. Thus, it can be assumed that thyroid autoimmunity involves these two antibodies equally. However, some of the patients and control subjects positive for antithyroglobulin antibody were not positive for antimicrosomal antibody, but none of the subjects were found positive for antimicrosomal antibody without being positive for antithyroglobulin. Many laboratories are now restricting their analysis of thyroid antibodies to the antimicrosomal antibody alone.2 23 If this is the case then a small percentage of patients with positive antithyroglobulin antibody will be missed, as has been demonstrated by this study. Therefore, it could be argued that antimicrosomal antibody is not entirely informative in the context of AITD.

About 9% patients would have been incompletely diagnosed unless antithyroid antibodies were assayed. Most of these patients were clinically suspected as having simple diffuse goitre, iodine deficiency goitre, multinodular goitre with hypothyroidism, and non-toxic solitary nodule. Out of the 36 mismatched patients, 33 had Hashimoto's thyroiditis. In the present study, calculated prevalence of autoimmunity among thyroid patients was 48.36%. Sensitivity of antibody tests in identifying AITD was not too high but the specificity was 87%. Thus tests for antithyroid antibodies can be considered useful in identifying existing thyroid autoimmunity which could not be easily predicted by clinical information only. This is important especially in multinodular hypothyroid, simple diffuse goitre, and iodine deficiency goitre, which are often considered clinically as iodine deficiency related disorders.

In conclusion, prevalence of AITD is quite high in Bangladesh. Antimicrosomal antibody is an important and specific marker for thyroid autoimmunity. An appreciable number of patients with simple diffuse, toxic multinodular, and iodine deficiency goitres were found positive for antithyroid antibodies indicative of coexistent autoimmune phenomenon, which may be missed unless they are investigated for antibodies.