Q J Med 2002; 95: 173-179
© 2002 Association of Physicians
Complement activation in postpartum thyroiditis
From the Departments of Medicine and 1 Medical Biochemistry, University of Wales College of Medicine, Cardiff and 2 Caerphilly District Miners' Hospital, Caerphilly, UK
Received 19 October 2001 and in revised form 12 December 2001
 |
Summary |
|---|
 Top Summary IntroductionMethodsResultsDiscussionReferences |
|---|
Background: Postpartum thyroid dysfunction (PPTD) develops in 50% of pregnant women who have raised levels of circulating thyroid peroxidase autoantibodies (TPOAb) at booking. Although these antibodies are able to activate the complement cascade in vitro, it is not known whether complement activation plays any role in the pathogenesis of this disease.
Aim: To investigate potential and actual activation of the complement system in women with postpartum thyroiditis.
Design: Complement activation was monitored on a weekly basis in 24 postpartum women who had raised TPOAb at 16 weeks gestation, attending an antenatal clinic in Mid-Glamorgan, Wales.
Methods: ELISA procedures were used to measure both in-vitro complement C3 activation by TPOAb and circulating terminal complement complexes (TCC) in serum.
Results: Higher levels of bioactive TPOAb activity were seen in women who developed PPTD when compared to those who did not. However, TCC remained undetectable in serum throughout the period of study.
Conclusions: In PPTD, despite the presence of circulating bioactive TPOAbs, the extent of complement activation is inadequate to cause detectable increases in peripheral blood TCC, suggesting that the complement system may not play a major role in PPTD pathogenesis.
|
Introduction |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
Postpartum thyroid dysfunction (PPTD) is a well-recognized cause of morbidity in the period following childbirth.1,2 Raised levels of circulating thyroid peroxidase autoantibody (TPOAb) are detected in 10% of pregnant women at 16 weeks gestation,3–5 of which 50% develop a destructive thyroiditis, characterized by transient hyperthyroidism and/or hypothyroidism during the first 6 months of the postpartum period.6 Permanent hypothyroidism is reported in as many as 30% of these cases after 3 years, and in 50% at 7–10 years.7
However, the pathogenesis of this common autoimmune endocrine problem remains elusive. Following the identification of thyroid peroxidase, the follicular membrane-bound enzyme, as the true antigen to the microsomal antibody, attention has focused on its possible pathological role in autoimmune thyroid disorders (AITD).8 To date, no clear mechanism has been defined for this autoantibody in AITD. The ability of TPOAb to fix complement and activate the complement cascade has been documented in Hashimoto's disease and Graves' disease.9,10 Further evidence for the involvement of complement in the pathogenesis of AITD is provided by immunohistochemical studies, which show terminal complement complexes (TCC) around the thyroid follicles of thyroidectomy specimens from AITD patients.11
These findings are supported by the detection of elevated levels of terminal complement complexes (TCC) in the serum of patients with Hashimoto's disease or Graves' disease.11,12 The last decade has witnessed a surge in interest in complement biology, with the discovery and cloning of complement regulatory proteins (CRP) and the development of sensitive assays for measuring both their activity in serum and the TCC formed as a result. An ELISA for TCC has been developed which specifically detects fluid-phase terminal activation complexes (Sc5b-9) in plasma.13 Also, Parkes et al. have developed an ELISA for the measurement of in vitro complement C3 activation by TPOAb.14 Using both assays, we have assessed complement activation in TPOAb-positive patients with or without PPTD. Unlike the majority of studies of this syndrome, where thyroid function and immunological changes have been monitored at monthly intervals or more, the patients in this study were monitored weekly in order that transient changes in complement activity and autoantibody levels might be better defined.
|
Methods |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
Patients
The patients described in this paper were recruited from women attending the antenatal clinic of Caerphilly District Miners' Hospital, Caerphilly, (CDMH), Mid-Glamorgan. Ethical approval for the study was obtained from the ethical committee of the Mid-Glamorgan Health Authority prior to commencement. Consecutive antenatal clinic attendees presenting at about 16–18 weeks gestation were screened for the presence of circulating thyroid autoantibodies (thyroid peroxidase and thyroglobulin autoantibodies). Of 468 women screened, 45 had significant levels of circulating autoantibodies to thyroid peroxidase (TPOAb
19.33 kIU/l) and were therefore at risk of developing PPT. Five patients with pre-existing thyroid disease and 16 patients who failed to complete the 24-week follow-up period were excluded from the study, leaving a total of 24 patients in the final group analysed. Weekly clinical assessments and venesections were performed between the 2nd and 24th weeks postpartum.
Laboratory studies
Thyroid function tests
Serum FT3, FT4 and TSH concentrations were all measured with Amerlite assay kits (Amersham International). The reference range for TSH was 0.5–3.6 mU/l, and its lower limit of detection was 0.04 mU/l.15 The normal reference ranges for FT3 and FT4 in these assays were 3–9 pmol/l and 8–26 pmol/l, respectively. The reference ranges were derived from the analysis of serum samples from 239 thyroid-antibody-negative women participating in the project. Hyperthyroidism was defined either as one or more occasions of elevated FT3 (>9 pmol/l) and FT4 (>26 pmol/l) or suppressed TSH together with elevated FT3 or FT4. Hypothyroidism was defined as either TSH >3.6 mU/l together with FT3 <3 pmol/l or FT4 <8 pmol/l or TSH >10 mU/l on one or more occasions.
Thyroid autoantibodies
Serum samples for antibody estimation were stored at -20 °C until completion of the patient's follow-up. Auto-antibodies to the thyroid microsomal antigen thyroid peroxidase (TPO) was measured by enzyme-linked immunosorbent assay (ELISA), based on the methods described by Groves et al.16 Thyroid microsomal fraction was prepared from snap-frozen Graves tissue by differential centrifugation.17 Ninety-six-well microtitre plates were coated with 100 µl per well of human thyroid ‘microsomes’ (10 µg protein/ml in 50 mM carbonate buffer at pH 9.3, containing 0.2 g/l sodium azide—coating buffer). The plates were incubated at 4 °C overnight and then washed four times with wash buffer (10 mM Tris (hydroxymethyl) methylamine in 150 mM sodium chloride, pH 7.4, containing 0.005% Tween 20). After washing, wells were filled with 1:100 dilutions of serum samples or standard dilutions in quadruplicate. Following incubation at room temperature for 2 h, the plates were washed as before. We then added 100 µl of peroxidase-conjugated sheep anti-human IgG antibody (Serotec) (diluted 1:2000 in 150 mM sodium chloride containing 1% sheep serum) to each well, and the plates were incubated at room temperature for a further 30 min. After washing as before, 100 µl peroxidase substrate (0.008% hydrogen peroxide and 0.4 g/l 2–2' Azino-bis(3-ethylbenzthiazoline 6-sulfonic acid) diluted in citrate phosphate buffer—anhydrous disodium hydrogen phosphate dissolved in 20 mM citric acid—at pH 5) was added to each well and colour was allowed to develop until the optical density (OD) of the highest standard reached approximately 0.9 OD units. Optical densities were measured in a Dynatech 500 Spectrophotometer at 492 nm.
The laboratory normal range for TPO antibody level (19.33 kIU/l) was determined was determined in a normal Welsh population as described previously.16
Terminal complement complex estimation
Plasma samples for measurement of TCC were stored at -70 °C until completion of follow-up. From the 24 women who comprised the final study group, 17 patients were selected at random and the samples were assayed in three batches. Of these, 11 developed PPTD during the period of weekly assessments, one developed PPTD later, four were antibody-positive but did not develop PPT and one patient had a pre-existing autoimmune thyroid disease.
TCC was quantified by ELISA, using a technique previously described by us.13 A mouse monoclonal antibody (obtained from Dr R. Wurzner, MRC Centre, Cambridge), WU7.2, was used as a capture antibody in this assay, to specifically bind fluid-phase terminal activation complexes (SC5b-9) from patient plasma samples. This antibody is specific for neo-epitopes of complement component C9 that are expressed only in the soluble TCC. Bound complexes were detected using a second antibody, a rabbit polyclonal antibody raised to human S-protein, another component of the soluble TCC. A standard sample for this assay was prepared from fully-activated normal human serum diluted 1:100 in phosphate-buffered saline (PBS+T) (10 mM sodium phosphate in 150 mM sodium chloride containing 0.1% Tween 20, at pH 7.3) and EDTA plasma, also diluted 1:100 in PBS+T. The EDTA plasma used had previously been defined as normal in this assay. A series of dilutions (10%–0%) were prepared from this standard and included in each assay. These standards were then used to construct a calibration graph for the assay.
Ninety-six-well microtitre plates (Dynatech) were coated with 100 µl per well of the mouse monoclonal antibody WU7.2 (1 µg/ml in 50 mM carbonate buffer at pH 9.6). After incubation at 4 °C for 24 h, the plates were washed once with PBS+T and 100 µl of 1% bovine serum albumin (BSA) in PBS+T added to each well. The plates were further incubated for an hour at 37 °C in order to block any non-specific binding of protein to the plate. The plates were washed three times with PBS+T, and 100 µl of plasma samples or standard dilutions (1:100 dilution) added to the wells in triplicate. The plates were then incubated for 1 h at 37 °C, to allow the capture antibody to bind any TCC present in the samples. After incubation and washing, 100 µl of the second antibody, rabbit anti-S protein IgG (diluted 1:2000 in PBS+T containing 1% BSA), was added to each well. The plates were again incubated at 37 °C for 1 h and washed three times as before; 100 µl of peroxidase conjugate (anti-rabbit IgG, horseradish-peroxidase-conjugated antibody, diluted 1:10000 in PBS+T containing 1% BSA) was then added to each well and incubation continued at 37 °C for another hour. The plates were then washed four times with PBS+T and 100 µl of OPD substrate solution (0.65 g/l 1,2-phenylene-diamine dihydrochloride and 0.01% hydrogen peroxide diluted in 0.1 M citric acid-phosphate buffer at pH 5.0) added to each well. Colour was allowed to develop for 5–20 min; the enzyme reaction was then stopped by adding 50 µl 1 M sulphuric acid to each well. The absorbances were read at 492 nm using a microplate spectrophotometer.
Complement C3 activation
Bound complement component C3b activated by TPO/TPOAb complexes was assayed in an ELISA using a peroxidase-conjugated goat anticomplement C3b antiserum.14 Microtitre plates were coated overnight with a human thyroid microsomal membrane fraction (10 µg/ml in 50 mM sodium carbonate, pH 9.1). Replicate wells were set up containing diluent alone as negative controls, NIBSC 66/387 reference preparation as primary standard, high C3 serum from an AITD patient as secondary standard or patient's serum (100 µl/well of 1:100 dilution) and the plates incubated at room temperature for 2 h. Complement-rich fresh guinea pig serum (Gibco, cat. no. 065-9202, stored frozen in aliquots at -70 °C), diluted 1:200 in assay buffer, was added to the wells, and the plates were incubated at room temperature for 1 h. After washing, the plates were treated with peroxidase-conjugated anti-guinea pig C3 antiserum (Cappel cat. no. 3207–0601, Dynatech). The degree of complement C3 activation was quantified following the addition of o-phenylenediamine (Sigma) and hydrogen peroxide as substrates in 50 mM citrate phosphate buffer (pH 5.1). The reaction was stopped after 3 min at room temperature by the addition of 50 µl of 2 M sulphuric acid, and the resulting colour was measured in a Dynatech 500 spectrophotometer at 419 nm.
C3 index was calculated from the optical density of the standard and test samples using the equation:
|
|
Statistical analysis
Data were analysed using the Student's t-test. Data for TPOAb and bioactive TPOAb activity were not normally distributed and required log transformation to achieve data skew values of 0±1. All antibody activities were expressed as geometric mean values with 95% after log transformation. The C3 index data did not require log transformation.
|
Results |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
Clinical
Of the 24 women studied, 13 developed one or more episodes of thyroid dysfunction during the postpartum period. Of these, eight showed a biphasic pattern of both hyperthyroidism (onset at 6–23 weeks postpartum) and hypothyroidism (onset at 14–28 weeks postpartum), three had hyperthyroidism alone (onsets at 9, 17 and 23 weeks) and three showed only hypothyroidism (onset at 15, 23 and 32 weeks postpartum). Persistent hypothyroidism was seen in three patients: two in the biphasic group and one in the hypothyroid PPT group (Table 1
).
|
TPOAb levels
Weekly TPOAb levels in the study population are shown in Figure 1. In the PPT group, the geometric mean TPOAb level was 188 kIU/l at 4 weeks, rising steadily to 566 kIU/l at 28 weeks. In the euthyroid antibody-positive group, the mean at 4 weeks was 125 kIU/l, but showed a less steep climb towards 28 weeks. The means at each time point did not differ significantly (p>0.05).
|
Complement C3 activation
The mean C3 index at 4 weeks postpartum was 0.23 in the PPT group (Figure 2). This increased steadily in the following weeks, peaking at 26 weeks (0.912) and then dropping slightly at 28 weeks. In the Ab-positive euthyroid group, the mean C3 index at 4 weeks was 0.06. This increased only minimally in the following weeks, attaining a maximum value of 0.195 at 26 weeks. The means differed significantly between 5–11 and 16–26 weeks postpartum (p<0.05).
|
Bioactive TPO autoantibody levels
Figure 3 shows the bioactive TPOAb levels in both women with PPT and euthyroid, Ab-positive women. The activity in both groups was low at 1 week post delivery (28.2 and 4.6 kIU/l, respectively). However, there was a rapid increase in the PPT group, with a peak of 342.8 kIU/l at 19 weeks and a small decline up to 23 weeks. In the euthyroid group there was only a minimal rise, with maximum values less than 30 IU/l. Significant differences between the two groups occurred between weeks 21 and 25.
|
Terminal complement complexes
In the TCC assays, the percentage activation in plasma recorded was <0.4% in all samples, as measured against the prepared standard curves. By this method, no significant complement activation was detected in any of the patient samples.
|
Discussion |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
Anti-TPO antibodies are a specific marker of PPTD but are unlikely to be the sole pathogenic agents for this syndrome. In this study, not all TPOAb-positive women developed PPTD, and postnatal TPOAb levels did not distinguish women who developed PPTD from those who did not. Rather, significantly higher TPOAb bioactivity was seen in the PPTD group when compared to the euthyroid antibody-positive group.
Significant differences in complement C3 activation were demonstrated from as early as 6–11 weeks, with a later period of significant difference at about 15–26 weeks. The bioactive anti TPOAb activity differed between the two groups, mostly from the 21st to 25th weeks postpartum. These periods do not appear to correspond to the clinically biphasic pattern of the illness. Clinically, median onset of hyperthyroidism was 12 weeks while median onset of hypothyroidism was 20 weeks. This suggests that there might be lag phases between clinical and immunological changes. Three women developed hypothyroidism, which was still present after follow up for 1 year. Their thyroid antibody levels and bioactive antibody activity fell within the same range as the other women who developed postpartum thyroiditis.
Though our numbers are small when compared to previous studies on TPOAb and complement activity in PPTD,17,18 ours is the first that looks at these parameters on a weekly basis. Follow-up of all positive TPOAb women at 16 weeks gestation was not achievable, but this is unlikely to have introduced serious bias, as the subjects who dropped out did so for varying reasons. While we confirm that women with significantly higher biological antibody activity are more likely to develop PPTD, we observe a different temporal pattern than that previously reported. Parkes et al. had reported highly significant differences between the two groups from about an earlier period of two months.17 This was not the case in this study, as the most significant difference in antibody biological activity between PPTD and non-PPTD antibody-positive women was observed from about 15 to 26 weeks postpartum.
Alongside other similarities, involvement of the complement system has been described in the pathogenesis of both Hashimoto's thyroiditis and PPTD.11,17 In this study, in-vitro complement activation by TPOAb was clearly demonstrable by ELISA, yet assays for TCC in plasma were consistently negative in all patients with PPTD. In Hashimoto's disease, bioactive TPOAb as well as TCC in plasma are consistently elevated.11
The reasons for these differences in TCC levels in the two conditions are not clear. These findings suggest that, in PPTD, complement-fixing TPOAbs are present in serum, but the extent to which they are able to activate the complement system is uncertain. Complement activation, if any, is certainly inadequate to cause a detectable increase in serum TCC levels. It is possible that complement regulatory factors may be involved in moderating complement activity in vivo. In patients with Hashimoto's and Graves' disease, the thyrocytes are relatively resistant to lysis by the homologous membrane attack complex due to the presence of certain MAC-inhibitory proteins such as CD59 antigen and membrane attack complex-inhibiting protein/homologous restriction factor (MIP/HRF) on these thyroid cells.19 In PPTD, expression of similar inhibitory proteins in serum may account for the absence of TCC in serum, and may provide important clues to the pathogenic differences between PPTD and Hashimoto's thyroiditis. On the other hand, TPOAbs may be unable to activate complement because the TPO-TPOAb antigen–antibody interaction necessary to trigger the cascade does not take place. Such a scenario is not unlikely, as TPO is normally expressed exclusively on the apical membrane of the thyroid cell where it is sequestered within the thyroid follicle. In the absence of significant thyroid damage, is unlikely that this autoantigen will come into contact with circulating antibodies. Finally, it is likely that other cellular factors, such as the rate of apoptosis, may be more important in thyroid cell destruction.20 Immunohistochemical examination of thyroid biopsies would be required provide clear evidence of complement activation in PPTD, but this is not feasible in this population of patients.
In conclusion, weekly monitoring of postpartum women showed that, while bioactive TPOAb activity in vitro was significantly higher in serum samples from women who develop PPTD, the absence of complement complexes in their serum suggests that complement activation is not a major factor in the pathogenesis of this autoimmune condition.
|
Notes |
|---|
Address correspondence to Professor J.H. Lazarus, Department of Medicine, University of Wales College of Medicine, Llandough Hospital, Cardiff. e-mail: Lazarus{at}cardiff.ac.uk
|
References |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
1. Amino N, Miyai K, Onishi T, Hashimoto T, Arai K. Transient hypothyroidism after delivery in autoimmune thyroiditis. J Clin Endocrinol Metab1976; 42:296–301.[Abstract]
2. Ginsberg J, Walfish PG. Postpartum transient thyrotoxicosis with painless thyroiditis. Lancet1977; 1:1125–8.[ISI][Medline]
3. Jansson R, Bernander S, Karlsson A, Levin K, Nillson G. Autoimmune thyroid dysfunction in the postpartum period. J Clin Endocrinol Metab1984; 58:681–7.[Abstract]
4. Lervang HH, Pryds O, Kristensen HPO. Thyroid dysfunction after delivery: incidence and clinical course. Acta Med Scand1987; 222:369–74.[ISI][Medline]
5. Fung HYM, Kologlu M, Collison K, John R, Richards CJ, Hall R, McGregor AM. Postpartum thyroid dysfunction in Mid-Glamorgan. Br Med J1988; 296:241–4.
6. Lazarus JH, Hall R, Othman S, Parkes AB, Richards CJ, McCulloch B, Harris B. The clinical spectrum of postpartum thyroid disease. Q J Med1996; 89:429–35.[Abstract]
7.
Premawardhana LDKE, Parkes AB, Ammari F, John R, Darke C, Adams H, Lazarus JH. Postpartum thyroiditis and long-term thyroid status: prognostic influence of thyroid peroxidase antibodies and ultrasound echogenicity. J Clin Endocrinol Metab2000; 85:71–5.
8. Czanocka B, Ruf J, Ferrand M, Carayon P, Lissitzky S. Purification of the human thyroid peroxidase and its identification as the microsomal antigen involved in autoimmune thyroid disease. Fed Eur Biochem Soc1985; 190:147–52.
9. Trotter WR, Belyavin G, Waddams A. Precipitating and complement-fixing antibodies in Hashimoto's disease. Proc Roy Soc Med1957; 50:961–8.
10. Khoury EL, Hammond L, Bottazo GF, Doniach D. Presence of organ specific microsomal autoantigen on the surface of human thyroid cells in culture: it's involvement in complement mediated cytotoxicity. Clin Exp Immunol1981; 45:316–28.[ISI][Medline]
11. Weetman AP, Cohen SB, Oleesky DA, Morgan BP. Terminal complement complexes and C1/C1 inhibitor complexes in autoimmune thyroid disease. Clin Exp Immunol1989; 77:25–30.[ISI][Medline]
12. Oleesky DA, Ratanachaiyavong S, Ludgate M, Morgan BP, Campbell AK, McGregor AM. Complement component C9 in Graves disease. Clin Endocrinol1986; 25:623–32.[Medline]
13. Morgan BP, Daniels RH, Williams BD. Measurement of terminal complement complexes in rheumatoid arthritis. Clin Exp Immunol1988; 73:473–8.[ISI][Medline]
14. Parkes AB, Harris R, Williams S, Lazarus JH, Waters JS. Elisa for the measurement of complement C3 activation by autoantibodies directed against thyroid membrane antigens. J Clin Lab Immunol1991; 35:139–45.[Medline]
15. John R, Henley R, Chang D, McGregor AM. Enhanced luminescence immunoassay: evaluation of new, more sensitive thyrotropin assay. Clin Chem1986; 32:2178–83.[Abstract]
16. Groves CJ, Howells RD, Williams S, Darke C, Parkes AB. Primary standardisation for the ELISA of serum thyroperoxidase and thyroglobulin antibodies and their prevalence in a normal Welsh population. J Clin Lab Immunol1990; 32:147–51.[Medline]
17. Parkes AB, Othman S, Hall R, John R, Richards CJ, Lazarus JH. The role of complement in the pathogenesis of postpartum thyroiditis. J Clin Endocrinol Metab1994; 79:395–400.[Abstract]
18. Parkes AB, Othman S, Hall R, John R, Lazarus JH. The role of complement in the pathogenesis of postpartum thyroiditis: relationship between complement activation and disease presentation and progression. Eur J Endocrinol1995; 133:210–15.[Abstract]
19. Tandon N, Morgan BP, Weetman AP. Expression and function of membrane attack complex inhibitory proteins on thyroid follicular cells. Immunology1992; 75:372–7.[ISI][Medline]
20. Borgerson KL, Bretz JD, Baker JR Jr. The role of fas-mediated apoptosis in thyroid autoimmune disease. Autoimmunity1999; 30:251–64[ISI][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||