Thyroid disease is relatively common in the primary care setting. This chapter reviews the basic approach to thyroid function testing, nonthyroidal illness, hyperthyroidism, hypothyroidism and thyroid nodules.

Assessment of Thyroid Function

Measurement of thyrotropin (TSH) concentrations and the free T4 index (FTI) are the best tests to evaluate patient suspected of having thyroid dysfunction. Figures 1 and 2 provide algorithms for evaluating thyroid function according to initial levels of TSH and FTI. Because of the log-linear relationship of TSH concentration to thyroid hormones, small changes in thyroid hormones result in large reciprocal changes in TSH. In most cases, a normal TSH value indicates the patient is euthyroid. However, there are several important caveats. TSH may be suppressed for two months following treatment of thyrotoxicosis when T4 and T3 levels are low and signs and symptoms of hypothyroidism are present. Similarly, TSH levels require up to two months to stabilize following a change in thyroid hormone dose. TSH levels are not a reliable measure of thyroid function in patients with central hypothyroidism due to hypothalamic or pituitary disease, TSH-producing pituitary adenomas, or thyroid hormone resistance. TSH secretion may be variably decreased by glucocorticoids (more than the equivalent of 20 mg/day prednisone), octreotide, dobutamine, and especially dopamine (more than 1 mcg/kg-min). On the other hand, TSH levels may be elevated in adrenal insufficiency, during recovery from illness, or patients receiving lithium.

Assessment of thyroid function is difficult during acute illness and recovery. Changes in levels of cytokines, thyroid binding proteins, and cortisol result in decreases in total concentration of thyroid hormones, inhibition of binding of thyroid hormones to binding proteins, decrease peripheral conversion of T4 to T3, and decrease in TSH secretion. This results in low serum levels of T4, T3, TSH, and variable FTI. The magnitude of changes in thyroid function correlates with the severity of the underlying illness. Values of FTI less than 3 have been associated with mortality rates of 85 percent. Almost all patients with detectable TSH (greater than 0.05 mU/L) will become euthyroid after recovery. However, during recovery, the levels of TSH may transiently rise above normal (up to 20 mU/L) while T4 and T3 are still low and resulting in an inappropriate diagnosis of primary hypothyroidism. Although this condition is commonly referred to as "euthyroid sick syndrome", there is evidence that this is a positive adaptive response to acute illness. Administration of thyroxine to critically ill patients has not been shown to alter mortality or outcome. Therefore, measurement of thyroid function should be deferred until complete recovery unless the patient has pre-existing thyroid disorder or there is a strong suspicion for primary thyroid disease.

A detailed medical history will usually reveal clues to the diagnosis of hyperthyroidism. Patients should be asked about nervousness, fatigue, palpitations, dyspnea, weight loss, heat intolerance, irritability, tremor, muscle weakness, sleep disturbance or change in menstrual pattern, bowel function, appetite, neck size or vision. Iodine exposure (including contrast agents and topical preparations), thyroid hormone use and family history of thyroid disease should also be considered. Symptoms of hyperthyroidism in older patients may be predominantly cardiac - angina, palpitations or congestive heart failure - or may be entirely absent.

When hyperthyroidism is suspected, measurement of serum thyrotropin (TSH) and FTI should confirm the diagnosis. If the TSH level is low but the thyroxine level is normal, serum triiodothyronine (T3) should be measured. Isolated elevations in T3 with normal FTI are present in up to15 percent of patients with hyperthyroidism- so-called T3 toxicosis. Total T4 concentrations may be increased in patients with high levels of thyroxine-binding globulin (TBG) but FTI and TSH will be normal. Rarely, hereditary mutations in albumin will cause high levels of T4 and FTI, but TSH will be normal.

The most common cause of hyperthyroidism is Graves’ disease. Other common causes include multinodular goiter, toxic thyroid adenoma and thyroiditis (which may be painful or painless). In the absence of typical physical findings of Graves’ disease - diffuse goiter, ophthalmopathy or pretibial myxedema – a radioiodine uptake scan should be performed. Low uptake values are indicative of thyroiditis, exogenous thyroid hormone, or iodine-induced thyrotoxicosis (including amiodarone).

Medical therapy of hyperthyroidism is summarized in Table 1. Cardiovascular symptoms may be ameliorated by beta-blockers, which also decrease the peripheral conversion of T4 to T3. Patients with Graves’ disease may be treated with thionamide drugs until remission, or prior to treatment with radioiodine. Treatment is usually started with 20 to 30 mg of methimazole once a day or 50 to 100 mg of propylthiouracil (PTU) three times a day. A clinical response occurs in 2-4 weeks in over 90 percent of patients. The dose should be adjusted every four to six weeks to maintain normal thyroid secretion. The interval between follow-up visits can then be extended to three months. Serious side effects include agranulocytosis (particularly with PTU), hepatitis, and lupus-like syndrome, which mandate discontinuation of therapy. Patients may also experience pruritic skin rashes. In the absence of serious side effects, therapy is usually continued for 18 months and then re-evaluated. PTU may have a greater risk of hepatotoxicity but has the benefit of blocking T4 to T3 conversion. It is drug of choice during pregnancy because methimazole has been associated with a congenital scalp disorder, aplasia cutis, in infants of treated women. The major drawback of drug treatment is that only 30-40 percent of patients with Graves’ disease will maintain a remission once the treatment is stopped.

Radioiodine is preferred treatment by most endocrinologists because of its greater efficacy in maintaining remission. Initial improvement occurs in 4-8 weeks in 70 percent of patients. Approximately one third of patients have transient elevation of thyroid hormones due to radiation thyroiditis. Therefore, patients with severe symptoms or cardiovascular disease should receive thionamides for 4-8 weeks prior to treatment. Rarely, ophthalmopathy may worsen, especially in smokers. Permanent hypothyroidism develops in 10-30 percent of patients during the first 2 years and 5 percent per year thereafter. Radioiodine does not appear to increase the risk of subsequent cancer. However, it is contraindicated during pregnancy and breast-feeding. Pregnancy should be avoided for 4 months following treatment.

Subtotal thyroidectomy is indicated in patients with large, compressive goiter, concurrent cold nodules, or in patients who fail drug treatment and refuse radioiodine therapy. Prior to surgery, patients should be medically treated until euthyroid. Patients with prominent symptoms of hyperthyroidism should be treated with beta-adrenergic antagonists for symptomatic relief. In the presence of asthma, or other contraindications to beta-blockade, calcium channel blockers may be used.

Hyperthyroidism caused by nodular disease (toxic adenoma or toxic multinodular goiter) does not remit spontaneously. The most appropriate therapies are radioiodine ablation or surgery. Hyperthyroidism associated with thyroiditis, which is common after pregnancy, is usually mild and temporary, requiring only observation or symptomatic treatment with a beta-adrenergic antagonist. If thyroid pain and tenderness are present, NSAIDS or glucocorticoid therapy may be added.

Low levels of TSH with normal values of FTI and T3 and no symptoms of hyperthyroidism are known as subclinical hyperthyroidism. Like subclinical hypothyroidism (see below), this condition is more common in the geriatric population. In a study from the UK, undetectable concentrations of TSH were present in 1.6 percent of the population over 60 years. The progression to overt hyperthyroidism in patients with multinodular goiter or autonomous adenoma is 5 percent per year. This risk is increased by exposure to iodine or the use of amiodarone. In contrast, the progression to symptomatic hyperthyroidism is probably less likely in patients with Graves’ disease. While management is controversial, there is evidence that untreated subclinical hyperthyroidism may have adverse effects on the cardiovascular system and skeleton. A cohort study of subjects over the age of 60 with TSH concentration less than 0.1 mU/L showed a 3.1 relative risk of developing atrial fibrillation over 10 years as compared with those with normal TSH and an absolute risk 28 percent. In two studies of subclinical hyperthyroidism, there was a clinically significant lower bone density in the femoral neck and radius. Two other studies have shown that treating postmenopausal women with subclinical hyperthyroidism reverses the 2 percent loss of bone density per year.

TABLE 1: Medical treatment of hyperthyroidism ^1,

The majority of cases of hypothyroidism are due to primary thyroid gland failure resulting from chronic autoimmune thyroiditis (Hashimoto’s disease), radioactive iodine therapy or surgery. Patients should be asked about symptoms of fatigue, weakness, sleepiness, cold intolerance, constipation, muscle cramps, mental impairment, depression, menstrual disturbances, infertility and weight gain (due to water retention). Clinical findings suggestive of hypothyroidism include goiter, bradycardia, edema, hoarseness, delayed relaxation of deep tendon reflexes, slow speech and cool, dry skin, and muscle tenderness.

To establish the diagnosis of hypothyroidism, a serum TSH measurement and a free T4 estimate should be performed. If the TSH is low, normal or insufficiently elevated in the presence of low T4 values, central hypothyroidism should be excluded.

Thyroxine is the drug of choice for hypothyroidism. Recent studies have shown that some brands of generic thyroxine are bioequivalent to the brand-name product, although some endocrinologists continue to recommend Synthroid. Adults require approximately 1.7 micrograms/kilogram a day and are usually started on 50 micrograms with dose escalation every two months. Patients over the age of 50, or younger patients with cardiac disease, should be started on a lower initial dose of 25 micrograms. Clinical and biochemical evaluation should be performed every six to eight weeks until the TSH concentration is normalized, and thereafter at six to 12 month intervals. If a change in dosage is required, the TSH concentration should be re-measured after two months. Patients with central hypothyroidism must also be evaluated for secondary adrenal insufficiency since administration of thyroid hormone increases the catabolism of cortisol and may precipitate adrenal crisis.

Five to ten percent of adult women, and as many as 15 percent of all patients over the age of 65 have "subclinical hypothyroidism," or mild thyroid failure, defined as normal T4 and elevated TSH but no symptoms. Patients with autoimmune disease such as Type I diabetes or celiac disease, a history of neck irradiation, treated thyrotoxicosis, or who are receiving drugs such as amiodarone, lithium, and interferon have an increased risk. As strikingly, two percent of women and 0.5-1 percent of men over 70 have unsuspected but symptomatic hypothyroidism, with elevated TSH and low T4 levels. When treated, symptoms of this overt but undiagnosed hypothyroidism are clearly reversible. Overt hypothyroidism during pregnancy is associated with a high rate of first trimester miscarriage, congenital malformations, and perinatal mortality. However, even subclinical hypothyroidism may have adverse neuropsychological consequences. A study of 7-9 year-old children whose mothers had elevated TSH concentrations during pregnancy revealed a significant decrease in neuropsychological scores; 15 percent had IQ scores of 85 or lower, as compared with 5 percent in the group whose mothers had normal TSH values.

When considering the controversial topic of screening for thyroid dysfunction, it is important to distinguish the two different syndromes mentioned above: subclinical hypothyroidism and symptomatic but undiagnosed hypothyroidism. The latter is clearly worth finding – symptomatic patients who are treated with thyroid replacement usually feel better.^2, The utility of treating patients in the former category – those with abnormal laboratory tests but no symptoms – has not been convincingly proven. Although subclinical hypothyroidism has been shown to be an independent risk factor for atherosclerosis and myocardial infarction in older women, there have not been large randomized controlled trials to demonstrate prevention of hypothyroid symptoms, prevention of thyroid disease progression, or reduction in coronary heart disease (due to improved lipid profiles and decreased homocysteine levels), all possible but unproven benefits of treatment. The three small randomized studies that have been done have had conflicting results.^,, Two studies showed significant improvement in symptoms in up to 28 percent of patients treated. In the study that found no benefit, however, the treated group still had high normal TSH concentrations (4.6 uM/L) following therapy. The risks of treatment include osteoporosis: in post-menopausal women who are treated for subclinical hypothyroidism, thyroxine replacement prevents the bone-conserving effects of hormone replacement therapy. Some authors recommend treating all patients with subclinical hypothyroidism, but most concede that more data are required.

There is considerable controversy regarding sceening (Table 2). Current guidelines from the American College of Physicians conclude that there is insufficient evidence to recommend for or against routine screening, but that "office-based screening to detect overt thyroid dysfunction may be indicated in women older than 50 years of age." The guidelines also note that screening all women over 50 and treating all those with subclinical hypothyroidism would result in four million new lifetime prescriptions for thyroxine in the first year and 600,000 to 1 million per 5 years thereafter.

While screening (and treating) completely asymptomatic patients is controversial, thyroid function tests are recommended for older patients with a history of thyroid disease, other autoimmune diseases, depression, unexplained hyponatremia, cognitive dysfunction, or hyperlipidemia. Some patients with these findings in the setting of an elevated TSH will feel better with thyroxine replacement: for patients with one symptom, it is estimated that 25 percent will have a clinical response to thyroxine replacement. Elderly patients with asymptomatic hypothyroidism and positive thyroid antibodies (anti-TPO) or TSH levels greater than twice normal have a high frequency of progression to overt hypothyroidism and thyroid replacement is generally recommended for this group. The goal of therapy should be to normalize TSH levels; overtreatment runs the risk of precipitating angina or reducing bone mineral density. If treatment is not given to these patients, their thyroid function tests should be monitored every one to two years.

In the absence of conclusive data, we recommend following the ACP guidelines and considering routine periodic TSH testing in all women over 50 years. This testing should be done in the outpatient setting, as screening for thyroid disease in inpatients is clearly less accurate. If the TSH is undetectable or 10 mU/L or greater, free thyroxine levels should be tested as well. Women with overt hypothyroidism should be treated. Those with TSH higher than 4 mU/L, normal thyroxine levels, and no symptoms of hypothyroidism have subclinical hypothyroidism; it is reasonable to treat the subset with positive autoantibodies, or TSH levels more than twice normal. Treatment of women without these markers is left to the discretion of physician and patient.

Case-finding – testing patients with specific disorders for thyroid dysfunction – is also a reasonable approach. The prevalence of subclinical hypothyroidism in men with hyperlipidemia was 1.8 percent; among women with hyperlipidemia the prevalence was 7.6 percent. Other studies have shown a similar doubling of rates of hypothyroidism among markedly hyperlipidemic patients. Diabetic patients, as well, have higher rates of hypothyroidism; one survey showed a prevalence of 10.8 percent among diabetics in a general medicine practice.

Any thyroid disease can appear as one or more thyroid nodules. The differential diagnosis of a solitary nonfunctional thyroid nodule includes adenoma, carcinoma, cyst, or - very rarely - inflammatory thyroid disease or a developmental abnormality.

The only biochemical test that is routinely needed in this situation is serum TSH, used to identify patients with unsuspected thyrotoxicosis. Fine-needle aspiration is the next test of choice and leads to the selection of patients who require surgery. The main indications for surgery are malignant or indeterminate cytology, local symptoms or neck disfigurement. Iodine radionuclide scanning should be reserved for patients with indeterminate cytologic features or thyrotoxicosis. Benign nodules are treated with thyroxine or observation (Figure 4).

Fine needle aspiration biopsy

(a) insufficient ® repeat

(b) benign ® follow

(c) malignant ® surgery

(a) cold ® surgery

(b) warm ® surgery

(c) hot ® treat if TSH suppressed

We thank Dr. Robert McConnell for his helpful comments and suggestions.

Many drugs inhibit absorption of exogenously administered T4, or influence thyroid function by changing the production, secretion, transport or metabolism of T4 and T3: