Hypothyroidism is the most common endocrinopathy of the dog as a result of the breed-associated incidence of lymphocytic thyroiditis. Clear syndromes of congenital hypothyroidism are recognized in the neonatal horse, but the incidence and pathogenesis in the adult horse is less understood. Spontaneous hypothyroidism in the cat and other domestic species is most commonly associated with congenital disease or following thyroidectomy. Less commonly, hypothyroidism can be caused by iodine deficiency or by the ingestion of goitrogenic substances in the environment or food; goitrogenic substances are compounds that interfere with thyroid hormone synthesis by the thyroid gland.

Clinical signs of hypothyroidism common to most species generally reflect the reduction in basal metabolic rate of the body and include lethargy, mental depression, weakness, inability to train, and/or nonpruritic hair loss (Ferguson, 1989a, 1993; Ferguson and Hoenig, 1991b; Ferguson et al., 1992; Peterson and Ferguson, 1990; Feldman and Nelson, 2004a).

Hyperthyroidism is now the most common endocrine disorder in the cat, and is only occasionally seen in other domestic species. Hyperthyroidism or thyrotoxicosis is caused by excessive concentrations of the circulating thyroid hormones, thyroxine (T₄) and triiodothyronine (T₃), most commonly the result of hyperplastic or benign adenomatous malignant thyroid glands in cats and adenocarcinomas in dogs. Hyperthyroidism occurs most frequently in middle-aged to geriatric cats and discussion of therapeutic agents will focus on those used in this species. The most common clinical signs associated with hyperthyroidism, which can be directly related to thyroid hormone excess, are weight loss in spite of ravenous appetite, hyperactivity, polydipsia, polyuria, diarrhea, intermittent fever, vomiting, and symptoms of cardiovascular disease such as tachycardia and dyspnea. Often the cats shed excessive amounts of hair or the coat may be matted. Rarely, hyperthyroid cats present in a way similar to what has been called “apathetic hyperthyroidism”; the cats are lethargic and often anorectic, perhaps representing an end-stage form of the disease (Feldman and Nelson, 2004b; Ferguson and Hoenig, 1991a; Peterson and Ferguson, 1990). Therapy focuses upon either definitive treatments such as radioiodide and thyroidectomy, or, as discussed below, the administration of drugs that reduce the synthesis, secretion, metabolism, and/or action of thyroid hormone.

Thyroid hormones are the only iodinated organic compounds in the body with the two major secretory products of the thyroid gland, thyroxine (L-T₄) and 3,5,3′-triiodothyronine (L-T₃) containing 65% and 59% iodine, respectively. The minimum iodine requirement of most animals is unknown, but the daily amount needed in the ration to prevent goiter in all animals is generally accepted to be about 1 µg/kg body weight. The daily recommended amount of iodine in the dog is 15 µg/kg and while it has not been carefully studied in the cat, it is believed to be about 100 µg/cat/day. Commercial cat foods have been shown to be quite variable in the amount of iodide provided, probably a result of the variable amount of seafood used to create the diet. Indeed, dietary iodide has been shown to be inversely related to free T₄ concentrations (Tarttellin and Ford, 1994). Also, a review has questioned the adequacy of iodide in commercial cat diets before 2006, suggesting that the increased incidence of feline hyperthyroidism might have been associated with iodide deficiency (Edinboro et al., 2010). Although true nutrient requirements for this micronutrient are not well established, most commercial dog and cat food preparations include at least three to five times the minimum stated requirement for iodine when fed in recommended amounts. As a result, today, documented iodine deficiency has become a rare condition in most domestic animals. During pregnancy, the recommended minimum daily requirement for iodine is increases fourfold. Areas of iodine deficiency in North America include the Great Lakes region and eastern British Columbia (Kaptein et al., 1994; Peterson and Ferguson, 1990; Zimmerman, 2013).

In humans, and increasingly recognized in domestic animals, a wide range of clinical conditions, such as chronic starvation or malnutrition, surgery, diabetes mellitus, hepatic and renal disease, and chronic systemic illness, may result in decreased serum T₃ concentrations together with elevated serum rT₃ values (Araujo and Carvalho, 2011). This “low T₃” syndrome results from inhibition of 5′-deiodinase, the enzyme necessary for conversion of T₄ to T₃ and the conversion of rT₃ to 3,3′-T₂ (Braverman and Utiger, 1991; Burrow et al., 1989; Kaptein et al., 1994; Kaptein, 1986; Daminet and Ferguson, 2003).

The reduction in the production of T₃, the most potent thyroid hormone, appears to be a beneficial adaptive mechanism by which the body serves to limit the loss of protein and perhaps blunt the metabolic rate during illness. Maintenance of serum T₃ concentrations (with triiodothyronine replacement therapy) in euthyroid fasting humans leads to excessive nitrogen excretion and blunting of the pituitary’s TSH response to TRH, as if there was a state of hyperthyroidism. The body of evidence does not support the contention that lowering of serum T₃ during malnutrition and illness is associated with “tissue” hypothyroidism. Also, outside of these regulatory mechanisms, there is, as yet, no evidence that a congenital or acquired 5′-deiodinase deficiency exists in animals or in specific tissues (Braverman and Utiger, 1991; Ferguson, 1984, 1988; Kaptein et al., 1993, 1994; Kaptein, 1986).

In acute and severe illnesses in humans, serum T₄ and T₃ concentrations may also fall, in what is called the “low T₄ state of medical illness.” Impaired serum protein binding of T₄ caused by inhibitors of binding (such as free fatty acids) or a reduction in binding protein concentration results in reduced total serum T₄ concentrations and increased free fractions of T₄. In most cases, however, the absolute free T₄ concentrations remain normal. A fall in serum TSH concentrations may also contribute to the subnormal serum T₄ concentrations, especially in human patients treated with dopamine or glucocorticoids, drugs inhibiting TSH release. No studies have examined systematically the benefit or detriment of thyroid hormone therapy in domestic animals; however, studies of critically ill patients with low serum T₄ and T₃ concentrations have revealed that thyroxine therapy is not beneficial and fails to improve survival (Kaptein, 1986; Kaptein et al., 1993, 1994).

The effects of nonthyroidal illness on thyroid hormone metabolism in the dog are less well characterized than in humans. In the dog, depressed serum T₄ concentrations have been reported in various nonthyroidal illnesses such as hyperadrenocorticism (e.g., Cushing’s syndrome), diabetes mellitus, hypoadrenocorticism (e.g., Addison’s disease), chronic renal failure, hepatic disease, as well as a variety of other critical medical illnesses requiring intensive care (Ferguson, 1984, 1988, 1994, 2007; Ferguson and Peterson, 1992).

This topic was reviewed by Daminet and Ferguson in 2003. A variety of drugs may impair plasma or tissue binding of the thyroid hormones or alter thyroid hormone metabolism. Drugs used in veterinary medicine that are most likely to alter circulating thyroid hormone concentrations include the glucocorticoids, anticonvulsants, quinidine, salicylates, phenylbutazone, and radiocontrast agents. The mechanisms by which these drugs exert their effect vary. Quinidine and other membrane-stabilizing drugs may inhibit 5′-deiodinase. Salicylates, furosemide, and oleic acid may directly displace thyroid hormone from plasma binding sites (Daminet and Ferguson, 2003; Ferguson, 2007; Ferguson et al., 2007). Phenylbutazone appears also to have a direct antithyroid (goitrogenic) effect in some species, having been shown to decrease total and free T₄ in the horse (Ramirez et al., 1997). In vitro, it appears to decrease serum hormone binding. Radiocontrast agents (e.g., diatrizoate, iopanoic acid, ipodate, tyropanoate, and metrizamide) may act by preventing the uptake of T₄ by tissue, by directly inhibiting 5′-deiodinase, or by releasing the iodine they contain to exert an antithyroid effect on the thyroid gland. No studies have been reported in domestic animals to evaluate the influence of these iodine-containing drugs on thyroid function tests or on subsequent radioiodine uptake (Ferguson, 1984, 1989b, 1994).

Exogenous glucocorticoids have also been shown to have a profound effect on thyroid function tests in the dog, but similar studies have had little effect on serum T₄ levels in the cat. A single high immunosuppressive dosage of glucocorticoid (2.2 mg/kg prednisone, 0.6 mg/kg dexamethasone) will lower serum T₃ but not serum T₄ concentrations in the dog (Kemppainen et al., 1983; Laurberg and Boye, 1984). Serum T₃ concentrations may be decreased because of glucocorticoid inhibition of 5′-deiodinase or simply because of a reduced availability of plasma T₄, the substrate for the enzyme. Most dogs on chronic, high-dose, daily glucocorticoid therapy will have very low or undetectable serum T₄ concentrations, as well as subnormal serum T₃ values (Ferguson and Peterson, 1992; Kaptein et al., 1992; Kemppainen et al., 1983; Moore et al., 1993; Torres et al., 1991). Based upon electron microscopic examination of thyroid tissue, it was postulated that glucocorticoids may interfere with thyroid hormone secretion by inhibiting lysosomal hydrolysis of colloid in the follicular cell (Woltz et al., 1983). Immunosuppressive dosages (1.1–2 mg/kg q 12 h) of prednisolone in dogs for 3 to 4 weeks significantly decreased serum total T₄ and to a lesser extent, FT₄ concentrations. These changes were observed as soon as 1 day after initiation of treatment (Torres et al., 1991; Daminet et al., 1999). Endogenous TSH was not affected by 3 weeks of an immunosuppressive dosage of prednisone, but the assay may not allow detection of a mild decrease. No changes in serum total T₄ levels were seen after a month of prednisone administration at an antiinflammatory dosage (0.5 mg/kg q 12 h) (Moore et al., 1993). However, a decrease in total T₃ concentrations with a supranormal T₄ increase after TSH administration was observed. Another study also showed that prednisone at an antiinflammatory dosage of 1 mg/kg/day had no effect on TSH concentration of dogs with experimental hypothyroidism, but higher dosages have not been studied. A subsequent study in T₄-supplemented hypothyroid dogs showed that prednisone at a dosage of 1 mg/kg PO q 24 h for 7 days decreased total T₄, while FT₄ was unchanged, suggesting that FT₄ may be less affected by daily prednisone administration. Subsequent antiinflammatory doses of prednisone administered every other day did not alter T₄ or TSH concentrations (O’Neill et al., 2010). In summary, glucocorticoids can markedly decrease total T₃ and T₄ concentrations and to a lesser extent FT₄ concentrations in dogs. Therefore, thyroid function test results should be interpreted carefully in any dog receiving glucocorticoids if prescribed an immunosuppressive dosage or for a period greater than 1 month (Ferguson, 1984, 1994; Moore et al., 1993; Daminet and Ferguson, 2003). In the horse, 50 g dexamethasone q 12 h topically for 10 days led to a 30–40% decrease in both T₃ and T₄ with nadir at day 2 for serum T₃ concentration and day 6 for serum T₄ concentration. Neither hormone concentration had recovered by 11 days after discontinuation of therapy (Abraham et al., 2011).

Other drugs have been well documented to alter thyroid hormone metabolism or serum or tissue binding of the thyroid hormones in the dog. The anticonvulsants diphenylhydantoin and phenobarbital, which are mixed function oxidase inducers, consistently decrease serum T₄ concentrations. Phenobarbital increases the rate of clearance of T₄. Increased hepatic deiodination of thyroid hormones, biliary clearance, and fecal excretion result in decreased concentrations of circulating thyroid hormones. The short-term administration (3 weeks) of phenobarbital to dogs did not affect total or free T₄ nor TSH serum concentrations in beagle dogs. However, several studies have looked at the long-term effects of phenobarbital administration. Total and free T₄ can be decreased to a range consistent with hypothyroidism. Endogenous TSH concentrations can remain within reference range or be slightly increased. Serum phenobarbital concentrations did correlate with the decrease in total and free T₄ concentrations in one study, but not in another. The reduction of free T₄ concentration particularly supports the idea that phenobarbital influences the steady-state clearance of T₄. Difficulty of interpretation of drug effects is compounded by the observation in dogs that seizure activity reduces serum total T₄ concentrations in proportion to the frequency of seizure episodes (von Klopmann et al., 2006). In one study, thyroid function normalized 1 to 4 weeks after discontinuation of phenobarbital. It is therefore recommended, when this is an option clinically, to evaluate thyroid function at least 4–6 weeks after discontinuation of phenobarbital (Kantrowitz et al., 1999; Gaskill et al., 1999; Müller et al., 2000; Gieger et al., 2000; Daminet et al., 1999; McClain et al., 1989; Curran and DeGroot, 1991; Johnson et al., 1993; Barter and Klaassen, 1994; Theodoropoulos and Zolman, 1989; Liu et al., 1995; DeSandro et al., 1991; Attia and Aref, 1991).

The antiepileptic potassium bromide is a halide chemically related to iodide and could potentially interact with iodine in the thyroid gland. In rats, bromide produces a relative iodine deficiency, thereby interfering with iodine uptake, iodine transport, and iodination of tyrosine and tyrosyl residues on thyroglobulin. In one study of epileptic dogs and another of healthy dogs receiving KBr at therapeutic dosages up to 6 months, no abnormal thyroid function test results were seen (Kantrowitz et al., 1999, Paull et al., 2000).

A study evaluated the effect of a standard dosage of trimethoprim/ sulfamethoxazole on thyroid function tests in dogs with pyoderma and normal baseline serum thyroxine concentrations. The average serum T₄, but not T₃, concentration fell significantly during the treatment period of 6 weeks. The TSH response also fell in several dogs and radionuclide imaging suggested that the preparation (likely the sulfa component) interfered with iodine metabolism by the thyroid gland. The thyroid carcinogenic potential of sulfonamides, which has resulted in restriction of some forms in food animals, is likely due to its goitrogenic potential, chronic elevation of serum TSH, and subsequent stimulation of thyroid growth (Hall et al., 1993). Sulfonamides are known goitrogens in domestic species because they can markedly interfere with thyroid hormone synthesis through reversible inhibition of TPO activity and reduction of serum concentrations of thyroid hormones. With reduced negative feedback, there is an increase in pituitary secretion of TSH, which induces proliferative changes in the thyroid gland. Long-term administration of sulfonamides and prolonged stimulation of the thyroid gland by TSH have been associated with thyroid neoplasia in rats. There is considerable interspecies variation with respect to TPO inhibition by sulfonamides. For example, only mild effects of sulfonamides are observed on human thyroid function. Several prospective studies have evaluated the effects of sulfonamide administration on canine thyroid function. Trimethoprim-sulfadiazine administered at a dosage of 15 mg/kg q 12 h to healthy dogs for 4 weeks had no effect on serum total T₄ or T₃ or free T₄ concentrations or upon the results of TSH stimulation tests. When trimethoprim-sulfadimethoxazole was administered to dogs with pyoderma for 6 weeks at a higher dosage of 30  mg/kg q 12 h, serum total T₄ and T₃, and FT₄ concentrations decreased. Half of the dogs in this experiment had total thyroxine (TT₄) levels below the reference range at the end of the treatment, which could easily have resulted in the inappropriate diagnosis of primary thyroid failure. In another study, the same dosage of trimethoprim-sulfadimethoxazole was given to healthy dogs for 6 weeks, and thyroid hormones were markedly decreased and endogenous TSH increased as soon as 7 days after initiation of the treatment. Thyroid imaging showed increased uptake of pertechnetate, and thyroid gland biopsies revealed hyperplasia of thyroid follicles and absence of colloid production. These findings further support the concept that sulfonamides act as goitrogens, resulting in primary hypothyroidism with secondary changes associated with the proliferative effects of TSH. The effects of sulfonamides on thyroid function are species specific, and dosage and duration dependent. Normalization of thyroid function test results in dogs after cessation of sulfonamide administration can take up to 8 or 12 weeks. Sulfa drugs can lead to clinical hypothyroidism and even goiter in some dogs (Cohen et al., 1980, 1981; Lagler et al., 1976; Panciera and Post, 1992; Post et al., 1993; Hall et al., 1993; Campbell et al., 1996; Gookin et al., 1999; Taeymans and O’Marra, 2009).

Many nonsteroidal antiinflammatory drugs (NSAIDs) have been shown to alter thyroid function tests in humans, since circulating thyroid hormones are highly protein bound, and various NSAIDs can displace thyroid hormones from serum protein-binding sites. Short-term administration of therapeutic dosages of salicylates leads to a transient increase in unbound hormone levels and suppression of TSH concentrations. After long-term treatment with salicylates, a new steady-state is reached, reflecting an increased T₄ turnover rate, and serum total T₄ concentrations have been reported to be reduced 20–40%. Free T₄ concentrations can be unchanged or decreased, and TSH concentrations return to the reference range within a few weeks of treatment. However, a 60-day double-blind study of osteoarthritic euthyroid dogs given placebo, meloxicam, carprofen, or chondroitin sulfate/glucosamine identified no drug-associated changes in serum total and free T₄ concentrations or TSH (Sauve et al., 2003).