Detection of autoantibodies against 3,5,3'-triiodothyronine in a hyperthyroid patient (a case report).
A 35-year-old female patient was evaluated for thyroid function and was found to be hyperthyroid clinically and by laboratory tests. During the course of treatment with neomercazole, she developed a bizarre picture with consistently low T3 levels irrespective of her clinical thyroid status and other laboratory tests like serum T4 and TSH levels. The serum of the patient when analysed for the presence of autoantibodies was positive for those against T3. The presence of T3 autoantibodies was confirmed by other laboratory techniques like assessment of blank values in radioimmunoassay, binding of T3-125I to isolated gamma globulin and starch gel electrophoresis. Autoantibodies were found to be present only against T3 and not against. T4 The affinity constant of T3 antibodies was 4 x 10(9) lit. mol-1 as determined by Scatchard plot analysis while total binding capacity was 23 x 10(-11) mol/L.
The most widely used investigations for assessment of thyroid function are estimations of total triiodothyro nine T3 and total thyroxine T4 in serum samples for which radio immunoassay is the method of choice. Although these tests are quite dependable for making correct diagnosis in majority of cases, they are not without drawbacks and several factors are known to influence the results, which then may not reflect the exact status of the thyroid gland. Factors which can vitiate these estimations include binding abnormalities to the serum proteins such as alterations in thyroxine binding globulin capacity or concentrations, familial dysalbumincmic hyper-thyroxinemia, prealbuminemic associated hyper-thyroxinemia and presence of auto-antibodies against T4 or T310 It is known that circulating auto-antibodies to thyroid hormones are found in various thyroid disorders . Their presence is usually suspected when inappropriate values are obtained for serum T4 or T3 by RIA. The fluctuations in the values obtained in such cases depend upon the method used for separation of bound and free antigens in RIA technique. We have investigated a patient who had auto-antibodies against T3 present in circulation and we describe here the methods used to confirm their presence.
A 35-year-old female patient attended the outpatient department of our centre for evaluation of thyroid function. She was found to be hyperthyroid both clinically and by laboratory investigations and was put on neomereazole two tablets thrice a day. She became hypo-thyroid while on therapy and on withdrawal of the medication, she showed a recurrence of thyrtoxicosis. During the course of follow-up she started showing low levels of serum T3 irrespective of her clinical status while the other standard tests of thyroid function like serum T4 and TSH correlated well with the clinical findings. Estimation of serum T3, T4 and TSH was performed using RIA kits supplied by Isopharm, Radio pharmaccutical Division, BARC, Mumbai. In view of the consistently low levels of T3, serum samples were analysed for the presence of antibodies, if any, against T3 and T4.
The patient did not dcmonstrate the presence of anti-thyroglobulin and anti-microsomal antibodies and did not suffer from any systcmic illness likely to influence the levels of thyroid hormones in blood. Concentration of thyroxine binding globulin in serum as determined by RIA (Cis International, France) was found to be 22.5 ug/ml, (19.6 ? 2.5). Serum albumin levels were within normal limits (3.8 g/dl, 4.1 ? 0.4). Serum prealbumin levels were not determined.
Detection of auto-antibodies in serum:
The presence of auto-antibodies against T3 or T4 was detected by performing the RIA procedure which included two sets of blank tubes. The first set of blank tubes were made using T3 free serum as a sample (assay blank) while the other had the patient's serum in its place (Patient's blank). The specific antibodies provided with the RIA kit were not added in any one of the blank tubes and were replaced by equal volume of buffer so that the total volume in all the tubes remains the same. Percentage of radioactivity obtained as bound fraction was taken as the non-specific binding. Since this binding was found to be higher while doing T3 RIA in this patient as compared to normal T3 free serum processed the same way, the presence of some binding abnormalities of thyroid hormones to carrier proteins or the presence of auto-antibodies against T3 was suspected ,. This was more likely to be due to presence of auto-antibodies since binding to carrier proteins was blocked by 8-anilino-1 naphthalene sulfonic acid (ANS) which was one of the ingredients supplied with the kit. Non-specific binding of T4-125 1 in T4 RIA was found to be identical to that in normal T4 free serum, hence auto-antibodies against T3 and not other binding abnormalities were thought to be present in the serum.
To confirm the presence of auto-antibodies against T4 following tests were done.
(1) Precipitation of ?- globulin: 0.5 ml of patients serum enriched with either T3-125 I or T4-125 I was treated with equal volume of 32% polyethylene glycol (PEG) carbowax-6000, which precipitated all ?- globulin . The mixture was centrifuged and supernatant was discarded. The pellet was washed with the same volume of 16% PEG solution and the precipitate dissolved in saline. Binding of T3-125 I and T4-125 I was measured and compared with those in normal pooled serum sample processed in similar way.
(2) Protein electrophoresis: Horizontal starch gel electrophoresis of the serum proteins was done by a conventional method in glycine-acetate buffer-pH 8.6. The serum samples were enriched with a tracer dose of high specific activity T3-125 I or T4-125 I with and without addition of ANS before subjecting them to electrophoresis. The gels were cut in one cm pieces and radioactivity was counted in a well type scintillation counter. The protein fractions were identified in relation to those seen in a separate portion of the gel in a normal pooled serum stained with Amido black.
Characteristics of the antibody:
Auto-antibodies precipitated by addition of PEG and solubilised in saline as described above were subjected to Scatchard plot analysis as described by Oliver and Cano6 so as to determine the affinity constant and total binding capacity for T3.
RIA of T3 was also done after extraction of the hormone in ethyl alcohol. The serum was treated with 95% ethanol which could precipitate most of the proteins including globulin and side by side extracted all the T3 associated with the carrier proteins as well as endogenous antibodies. Estimation of T3 in the supernate was done by RIA as described by Wener et al.
[Table - 1] shows the thyroid status of the patient, dose of anti-thyroid drug, serum levels of TSH, T3, T4 and respective blank values at different time intervals during the follow-up period. It is evident from the Table that the patient's serum samples gave T3 values lower than expected (33, 9, undetected and 43 mg/100 ml) together with high blank values (39, 48, 41 and 33%) in comparison with blank values given by normal T3 free serum. Elevated blank values indicated the presence of an additional binder(s) which could be auto-antibodies. This was confirmed by precipitating all the ?- globulin with PEG and measuring the binding of radioactive T3 to the ?- globulin after the pellet was dissolved in a saline solution. It can be seen in [Table - 1] that binding of T3-125 I to antibodies precipitated by PEG was as high as 42.4% of the added radioactivity in comparison with 8.6% in normal T3 free serum. The binding of isolated g-globulin from a normal pooled serum was found to be more or less similar to the blank values obtained in RIA procedure. On the other hand, serum T4 levels all throughout gave expected values and related well to the thyroid status of the patient. Also blank values in case of T4 in the patient sample were identical to those obtained in normal T4 free serum. This indicated that auto-antibodies against T4 were not present in the circulation.
The presence of auto-antibodies against T3 and absence of these against T4 in patient's serum was further confirmed by subjecting the serum sample to starch gel electrophoresis. The measurement of radioactivity in one cm segments of the gel revealed that the percentage T3-125 I distributed was 52.1% at TBG, 24.4% at albumin and prealbumin fractions and 21.9% at g-globulin fraction in case of patient's serum. In contrast to this, electrophoresis of normal pooled serum sample showed radioactive concentration of 65.7% at TBG and 33.5% at albumin and prealbumin fractions. There was no significant radioactivity in the region of g-globulin. When the patient serum sample was enriched with ANS, a compound used to block the binding of thyroid hormones to that carrier proteins  and when subjected to clectrophoretic procedure, it was seen that added T3-125 I concentrated mainly in ?-globulin fraction to an extent of 90%. The electrophoresis done on patient's serum as well as normal pooled serum sample after addition of T4-125 I gave identical distribution pattern of radioactivity with no concentration at ?-globulin fraction.
The affinity constant of the auto-antibodies against T3 as measured by Scatchard plot analysis was found to be 4 x 109 lit. mol-1 and total binding capacity was 23 X 10-11 mol/L.
Presently quite a few reports are available where the presence of auto-antibodies which bind T4 or T3 or both in serum have been described,,,,,. The presence of antibodies is suspected when irrelevant results of estimations of these hormones are obtained depending upon the method used for the separation of bound and free hormones in the RIA systcm. Use of coated charcoal as in B. A. R. C. T3-RIA kit or PEG precipitation method as in B. A. R. C. T4-RIA kit give values which are underestimated while coated tubes give elevated results ,,. It happens because of the interference in RIA due to the presence of endogenous antibodies which act as an additional binder. 'Our patient started giving abnormally low values of T3 even when she was hyperthyroid or eu-thyroid. The presence of antibodies against T3 was amply demonstrated and confirmed by various techniques described above. Labelled T3-125 I showed higher blank values revealing the presence of an additional binder which was proved to be ?- globulin on their selective precipitation by PEG and further confirmed by migration of radioactivity together with ?-globulin fraction in starch gel electrophoresis. Levels of thyroxine binding globulin and those of albumin in serum were normal. This suggested that alterations in concentrations of carrier proteins were not responsible for low T3 levels. These experiments also demonstrated that antibody against T4 was not present since this tracer did not show high blank values in RIA or any binding with y-globulin region in clectrophoresis.
A simple experiment thereby rcmoving endogenous antibodies and other proteins by treatment with 95% ethanol and estimation of T3 by RIA in the ethanol extract proved this point further. The concentration of T3 after this procedure was found to be 163 ng/100 ml with normal blank value (3.3%), while the same serum sample had given a level of 33 ng/100 ml with high blank value on direct RIA when the patient was eu-thyroid [Table - 1].
The hormonal level that is obtained after alcohol extraction cannot be considered always a true value indicative of the thyroid activity of the patient since that will depend upon how high or low is the titre of the auto-antibodies and the amounts of hormone associated with thcm. In such cases, free thyroid hormone level is the better indicator of the thyroid activity.
The auto-antibodies against T3 present in the serum of our patient had an affinity constant of 4x10 lit. mol-1, and their binding capacity was 23 X 10-11 mol/L. These results are more or less similar to those obtained by Staelieli et a1 who reported the affinity constant of T3 antibodies in his patient to be 5.4 x 108 lit. mol-1 and the binding capacity of 1.4 ng T3/ml serum. We were not able to find out the clonal characteristics of these auto-antibodies, however the shape of the Scatchard plot suggested that these were the antibodies with variable affinity constants hence they are likely to be of polyclonal nature. Auto-antibodies against T3 were found to be present in the human serum which were both either polyclonal or monoclonal.
The etiology of the endogenous production of the auto-antibodies against T3 or T4 is not clear. Some authors, postulated synthesis of such antibodies after external administration of desiccated thyroid as there is a possibility of absorption of fregments of thyro-globulin from G. I. Tract,. Our patient as well as those described by others were never on thyroid extract even then they demonstrated the presence of auto-antibodies against thyroid hormones. Hence, it appears that complex Immunological reactions are responsible for production of such antibodies. The incidence of presence of auto-antibodies against thyroid hormones was reported to be about 1% by Mulingar et al . Thus the occurrence of these auto-antibodies is not that common. It is essential that the potential risk of interference due to their presence while performing RIA of thyroid hormones by any systcm should be kept in mind by the physicians as well as laboratory personnel.