3.4.6 Thyroid disease after pregnancy: postpartum thyroiditis

Postpartum thyroiditis is defined as an exacerbation of autoimmune thyroiditis during the postpartum period (1). Patients do not develop thyroid autoimmunity at the onset of postpartum thyroiditis, but have ‘subclinical autoimmune thyroiditis’ beforehand which is exacerbated after delivery. Typically an exacerbation induces destructive thyrotoxicosis followed by transient hypothyroidism. However, various types of thyroid dysfunction may occur, including Graves’ disease. Therefore, any kind of thyroid dysfunction observed during the postpartum period, is referred to as ‘postpartum thyroid dysfunction’.

The pathogenesis of postpartum thyroiditis is similar to that of postpartum exacerbation of Graves’ disease or Hashimoto’s thyroiditis, which occurs by the enhancement of immune activities after parturition. The difference is only their stage of autoimmune disease. In postpartum thyroiditis, immune activation causes the transition of subclinical into overt autoimmune thyroid disease, whereas in previously manifest Graves’ or Hashimoto’s disease, immune activation results in exacerbation or relapse after parturition. During pregnancy, maternal immune activities are suppressed in order to prevent rejection of the fetus. Sudden release from the immune suppression at the time of delivery intensifies immune activities above the normal level, just as the sudden cessation of immunosuppressive drugs gives rise to the exacerbation of autoimmune diseases (2). The serial changes in titres of microsomal (thyroid peroxidase) antibodies in pregnant women with Graves’ disease and Hashimoto’s disease (Fig. 3.4.6.1) support this view (3). The immune rebound seems to be a general phenomenon observed in the postpartum period, since serum levels of immunoglobulins, and counts of lymphocytes and natural killer (NK)/K cell activity decrease in late pregnancy and increase after delivery even in normal pregnant women (2, 4). As immunological situations after abortion are similar to those during the postpartum period, postabortional thyroid dysfunction may occur in some cases (5, 6). The postpartum rebound of immune activities comprises two phases. Cytotoxic T cells and NK cells increase from 1 to 4 months postpartum (Fig. 3.4.6.2) (7, 8). The enhancement of cellular immunity may exacerbate tissue injury in Hashimoto’s thyroiditis. In contrast, CD5 B cells, which produce autoantibodies, increase from 7 to 10 months postpartum (Fig. 3.4.6.2) (7). The enhancement of humoral immunity may cause postpartum Graves’ disease by an increase of antithyroid-stimulating hormone (TSH)-receptor autoantibodies. Indeed, Hashimoto’s thyroiditis is commonly aggravated from 1 to 4 months postpartum and Graves’ disease may develop or relapse from 4 to 12 months postpartum (Fig. 3.4.6.3).

A recent study on the production of cytokines revealed that T helper 1 (Th1)-type and T helper 2 (Th2)-type cytokines decreased during pregnancy, Th1-type cytokines increased during the early postpartum period and Th2-type cytokines increased during the later postpartum period (9). These data also strongly support the immune rebound hypothesis (2). A possible role of fetal microchimerism was proposed as the mechanism of postpartum exacerbation of autoimmune thyroid disease (10), but it is necessary to accumulate more evidence to prove this hypothesis.

Postpartum thyroid dysfunction is a very common phenomenon (2, 11) with an incidence of about 5% (1.1–16.7%; Table 3.4.6.1) in mothers in the general population, i.e. one in 20 pregnant women develop postpartum thyroid dysfunction (11). Postpartum thyroid dysfunctions are classified into five groups by their clinical features, hyperthyroid and/or hypothyroid, transient or persistent (Fig. 3.4.6.4):

Patients with persistent thyrotoxicosis (group 1) and some with transient thyrotoxicosis (group 2) reveal a high radio-iodine uptake due to Graves’ disease. Transient thyrotoxicosis with a high radio-iodine uptake is common in postpartum Graves’ disease; the overproduction of thyroid hormones ceases spontaneously within a year. The prevalence of postpartum Graves’ disease (both persistent and transient) is estimated at 11% of those with postpartum thyroid dysfunction and 0.54% of the general population (12). Thyrotoxicosis due to postpartum Graves’ disease occurs between 3 and 10 months postpartum.

The other three types of postpartum thyroid dysfunction are associated with thyroid tissue damage due to an exacerbation of autoimmune thyroiditis. They often manifest themselves as transient thyrotoxicosis (destructive thyrotoxicosis) developing at 1–3 months postpartum. Depending on the extent of the destruction, transient hypothyroidism may follow (group 3) or not (group 2, with a low radio-iodine uptake).

Occasionally, Graves’ disease occurs closely following, or concomitantly with destructive thyrotoxicosis (13). When cellular damage occurs slowly, hypothyroidism alone, rather than destructive thyrotoxicosis, may be observed after delivery. In many cases, it is transient (group 4). However, it may be persistent in a few cases (group 5). Destructive transient thyrotoxicosis (group 2, with a low radio-iodine uptake, and group 3) is the most common form of postpartum thyroid dysfunction, accounting for 50–60% of all postpartum thyroid dysfunction. The rest (groups 4 and 5) show only the hypothyroid phase, however, persistent hypothyroidism is very rare (<0.1%).

Thyroid dysfunction is most often subclinical: the patient has no complaints of hyper- or hypothyroidism, and thyroid function tests reveal only mild changes in serum TSH and thyroid hormones. Symptoms and signs in overt postpartum thyroid dysfunction are no different from those in nonpostpartum cases. Hypermetabolic and hyperdynamic symptoms, such as palpitation, sweating, and finger tremor, can be observed in any type of postpartum thyrotoxicosis. In postpartum Graves’ disease, eye signs and/or pretibial myxoedema may be present. In postpartum hypothyroidism, symptoms such as weakness, fatigue, dry skin, constipation, and cold intolerance, and signs such as cold skin, bradycardia, and thyroid enlargement are common. Since the hypothyroidism is of short duration, there is little risk of myxoedema. Postpartum depression, sometimes found with postpartum thyroid dysfunction, is an important problem. Depression was found to be associated with antithyroid autoantibodies rather than hypothyroidism (14).

Diagnosis of postpartum thyroid dysfunction is simple when the patient shows abnormal thyroid function tests during the postpartum period and positive thyroid autoantibodies. Thyroid dysfunction, however, is most often subclinical. For patients with overt thyrotoxicosis, it is essential to differentiate between postpartum Graves’ disease and destructive thyrotoxicosis. Usually an educated guess can be made from the time of onset (4–8 months after parturition in Graves’ disease versus 1–3 months in destructive thyrotoxicosis). Anti-TSH-receptor antibody and other markers helpful for the differential diagnosis are summarized in Table 3.4.6.2 (15, 16). Blood tests, however, are not conclusive since anti-TSH-receptor autoantibodies are sometimes found in Hashimoto’s thyroiditis and other tests do not have distinct cut-off values. The quantitative measurement of thyroid blood flow using Doppler ultrasonography is useful for differentiation and it is high in Graves’ disease and low in destructive thyrotoxicosis (17). The measurement of radioactive iodine uptake gives a definitive clue in the differential diagnosis between Graves’ disease (high uptake) and destructive thyrotoxicosis (low uptake). However, it should not be performed in mothers who are breastfeeding. There is no reliable way to differentiate between transient and persistent Graves’ disease.

For hypothyroidism, most cases are transient and due to an exacerbation of autoimmune thyroiditis. The finding of positive antithyroid microsomal antibody and/or antithyroglobulin antibody supports the existence of autoimmune thyroiditis, but negative results are obtained in 5–30% of patients. Cases of iodine-deficient hypothyroidism may occur in areas where iodine intake is marginal or mildly deficient, such as Europe (18), although it is likely that mothers already had hypothyroidism during pregnancy. Once the diagnosis is established, patients should be followed up for 1 year since Graves’ disease may occur shortly after destructive thyrotoxicosis.

In postpartum thyroid dysfunction some immunological abnormalities are observed before the onset of thyroid dysfunction (and, therefore, before and during pregnancy). Among these, the measurement of thyroid peroxidase antibodies or microsomal antibodies (MCAb) is the most useful marker for the prediction of the occurrence of postpartum thyroid dysfunction (19). If thyroid peroxidase antibodies are present, there is always lymphocytic infiltration into the thyroid and, therefore, ‘subclinical autoimmune thyroiditis’(20), which may be exacerbated after delivery. Of women with a positive measurement of thyroid peroxidase antibodies in early pregnancy, 60–70% develop postpartum thyroid dysfunction (2), whereas the risk of developing postpartum thyroid dysfunction in women with negative thyroid peroxidase antibody is estimated to be 0.6%. Mothers with a high thyroid peroxidase antibody titre (MCAb more than 5000–10 000 reciprocal dilution) always develop postpartum thyroid dysfunction. However, the measurement of thyroid peroxidase antibodies does not provide any information on the type of dysfunction that will occur. Although the measurement of thyroid peroxidase antibodies with semiquantitative antimicrosomal particle agglutination (MCPA) tests is simple and cheap, the value of screening for postpartum autoimmune thyroid syndrome remains unresolved (see also Chapter 3.1.7), probably depending on each country’s national health system (21). Antithyroid peroxidase antibody is as useful as semiquantitative particle agglutination tests for predicting postpartum thyroid dysfunction (22).

Graves’ disease is aggravated in early pregnancy, ameliorates in the latter half of pregnancy, but often relapses postpartum (23). Human chorionic gonadotropin (hCG) plays a crucial role in the aggravation of Graves’ thyrotoxicosis in early pregnancy (24). The relapse of Graves’ thyrotoxicosis after parturition may occur even in patients in remission before pregnancy. The new onset of Graves’ disease in the postpartum period is of great interest, since early diagnosis and treatment at the onset of the disease can lead to early remission (12). It is also an important period of Graves’ onset, since 40% of Graves’ patients who have had one or more deliveries developed their disease postpartum (25). TSH-receptor stimulating antibodies can be taken as a marker for postpartum development of Graves’ disease, since TSH-receptor stimulating antibodies are positive before the onset of Graves’ disease (26) when measured with a sensitive bioassay. Pregnant women with positive TSH-receptor stimulating antibodies in early pregnancy have a high-risk of developing postpartum Graves’ disease. Seventy-one pregnant women with positive thyroid peroxidase antibodies in early pregnancy were prospectively observed from early pregnancy to the postpartum period (12). Among them, seven showed positive TSH-receptor stimulating antibodies, five (70%) of whom developed postpartum Graves’ disease. Thyrotoxicosis in three of those five was transient and spontaneously improved within a year. Graves’ disease did not occur in the TSH-receptor stimulating antibody-negative patients (Fig. 3.4.6.5). The conventional radioreceptor assay for anti-TSH-receptor antibodies was not able to discriminate postpartum Graves’ disease.

In postpartum Graves’ disease, treatment options are antithyroid agents, radioactive iodine, or subtotal thyroidectomy, as in ‘typical’ Graves’ disease. Antithyroid drug treatment is a good initial choice because: (1) postpartum Graves’ hyperthyroidism is often transient, (2) even in persistent Graves’ disease the early diagnosed patients are easily controlled with antithyroid drugs (12), and (3) mothers may not want breastfeeding to be interrupted by radio-iodine therapy (see also Chapters 3.3.6 and 3.3.9). Radioactive iodine treatment can still be applied when hyperthyroidism persists after 1 year and has not gone into remission during antithyroid drug treatment.

In destructive thyrotoxicosis, the thyrotoxic phase is always transient and spontaneously ceases in 1–3 months. Treatments should be symptomatic, mainly with β-adrenergic antagonists for cardiovascular hyperdynamic symptoms. Because breastfeeding should be discontinued, β-adrenergic antagonists are indicated only in severe cases. Antithyroid therapy is not indicated. The treatment of hypothyroidism is required only when the patient has symptoms of hypothyroidism. Usually thyroxine (T₄) therapy with a gradual reduction in dose may go well, but recovery of patient’s thyroid function cannot be followed explicitly. Alternatively, replacement with a submaximal dose of triiodothyronine (15–50 μg T₃ daily by mouth in one to three divided doses) is useful in most transient cases, since spontaneous recovery of thyroid function can be monitored by an increase of serum T₄. In permanent hypothyroidism, when serum T₄ does not recover after several months of T₃ treatment, T₄ replacement is indicated. Recently successful prevention for postpartum development of thyroid dysfunction was achieved by short-term immunosuppressive therapy in patients who were predicted to develop postpartum hypothyroidism (27). It is also reported that selenium supplementation during pregnancy and in the postpartum period reduced thyroid inflammatory activity and the incidence of hypothyroidism (28).

Little is known about the long-term prognosis of postpartum Graves’ hyperthyroidism, although a better outcome than in ‘typical’ Graves’ disease might be expected in view of early diagnosis just after the onset of disease. In destructive thyrotoxicosis and/or hypothyroidism due to exacerbation of autoimmune thyroiditis, thyroid dysfunction is transient and most patients recover spontaneously to euthyroidism. Only in a few cases, hypothyroidism may persist. High titres of thyroglobulin antibodies and/or thyroid peroxidase antibodies are risk factors of persistent hypothyroidism. Even after recovery from hypothyroidism, abnormalities in ultrasonography and/or iodide perchlorate discharge tests may persist for a long time (29, 30), reflecting underlying chronic autoimmune thyroiditis. The patients almost certainly will develop postpartum thyroid dysfunction after the next parturition, with similar time of onset, type of thyroid dysfunction, and duration of dysfunction as in the previous episode.

Late development (after 5 years or more) of permanent hypothyroidism is found in 25–60% of the patients with postpartum thyroiditis (31–33) and, therefore, these patients should be followed up at appropriate intervals (once every 1–2 years) (34). Othman et al. reported that high titres of anti-MCAb and the severity of the hypothyroid phase of postpartum thyroiditis are risk factors for the late development of permanent hypothyroidism, but there was no association with HLA haplotype or family history of thyroid disease (31). In contrast, in Japanese women, high titres of thyroglobulin antibodies and HLA-DRw9 and/or B51 genotype were risk factors of permanent hypothyroidism (32).