3.5.2 Management of nontoxic multinodular goitre

Goitres can be classified according to thyroid function into toxic goitres, hypothyroid goitres, and euthyroid or nontoxic goitres (see Chapter 3.5.1). The most prevalent causes of nontoxic goitre are endemic (iodine-deficient) goitre and sporadic nontoxic goitre (diffuse or nodular). The disease entity of sporadic nontoxic goitre is defined as a benign enlargement of the thyroid gland of unknown cause, in euthyroid patients (normal serum free thyroxine (T₄) and free triiodothyronine (T₃) concentrations) living in an area without endemic goitre. The diagnosis is by exclusion. The prevalence of sporadic nontoxic goitre (also called simple goitre) in the adult population is high, 3.2% in the UK (see Chapter 3.1.7), and it is more common in women (5.3%) than in men (0.8%). This chapter deals predominantly with sporadic nontoxic multinodular goitre.

In a cross-sectional survey of 102 consecutive patients referred because of sporadic nontoxic goitre, goitre size is positively related to age and to duration of goitre (1) (Fig. 3.5.2.1). Patients with a multinodular goitre are older and have a larger goitre than patients with a diffuse or uninodular goitre. Plasma thyroid-stimulating hormone (TSH) is negatively related to goitre size (Fig. 3.5.2.2). Patients with a multinodular goitre and a suppressed TSH are older and have higher plasma free T₄ concentrations and larger goitres than those with a multinodular goitre and a normal TSH. The data suggest a continuous growth of nontoxic goitre and provide support for the concept of increasing thyroid nodularity and autonomy of thyroid function, related to increasing goitre size, during the natural history of the disease (see Chapter 3.5.1).

Several nontoxic goitre patients have no symptoms at all, or just complaints of cosmetic disfigurement (Box 3.5.2.1). Local discomfort in the neck is very common. Obstructive symptoms range from very slight to severe, caused by compression of neighbouring structures such as the upper airways, the recurrent laryngeal nerve, the oesophagus, and the great veins in the thoracic inlet. Not all compressive goitres, however, are symptomatic. Upper airway compression may cause dyspnoea, cough, and a mild choking sensation aggravated by recumbency, but these complaints occur only in one-half of the patients in whom tracheal compression is noted (2). The trachea presumably needs to lose 75% of its cross-sectional area before a stridor is clearly recognizable (3). The inspiratory stridor may be noticed on deep inspiration but not on ordinary breathing, and is frequently related to recumbency. In patients with substernal goitre, 8–13% complain about hoarseness. Vocal cord paresis occurs in 3–4% of substernal goitres, apparently due to stretching and ischaemia of the recurrent laryngeal nerve; it is not by itself indicative of malignant growth (4). Extrinsic pressure on the oesophagus produces dysphagia in about 20% of patients (2).

The presence of a goitre is usually ascertained by inspection and palpation of the neck. Agreement between observers on the presence of a goitre and on the diffuse or nodular nature of a goitre is low (5). Haemorrhage in a nodule may cause local neck pain for a few weeks. Some retrosternal extension of the goitre is quite common, but intrathoracic or substernal goitres (defined as having its greater mass inferior to the thoracic inlet) occur only in 3–5%, especially in elderly women with long-standing compressive goitres. The thyroid gland is not palpable in 10–30% of substernal goitres (4). Rarely, a ‘goitre plongeant’ or plunging goitre is observed: the goitre disappears into the thoracic cavity and reappears in the neck on swallowing or coughing. Clinical clues to the presence of a substernal goitre (in the absence of a cervical mass) are facial plethora, dilated veins over the thoracic inlet, and nocturnal dyspnoea when the patient sleeps on the side of the goitre. In this respect Pemberton’s sign is helpful too: the patient is asked to elevate both arms until they touch the sides of the face, and the presence of a substernal goitre narrowing the thoracic inlet and obstructing the great veins may reveal itself after a few moments by congestion of the face, some cyanosis, and distress. A goitre occluding the thoracic inlet has been named appropriately ‘the thyroid cork’ (3). Downhill oesophageal varices secondary to obstruction of the superior caval vein is rarely reported in substernal goitre.

Having completed the history and physical examination of the patient, determination of plasma TSH and thyroid ultrasonography and scintigraphy usually suffice for assessing the functional and anatomical characteristics of the goitre.

Subclinical hyperthyroidism (i.e. a suppressed TSH in the presence of a normal free T₄ and free T₃) is present in about 20% of patients, especially in the older age group with large multinodular goitres (1). It is prudent to determine plasma T₃ in these cases in order not to overlook T₃ toxicosis, which is not uncommon in multinodular goitres and requires treatment (see Chapter 3.3.11). Plasma thyroglobulin can be markedly elevated as it is positively related to goitre size (1), but its determination serves no useful purpose. Serum thyroid peroxidase antibodies are found in 15–20% of patients; if present, the patient is at risk of developing Graves’-like hyperthyroidism or hypothyroidism after ¹³¹I therapy (see below).

Radionuclide scintigraphy with ¹²³I or ^99mTc pertechnetate usually visualizes the goitre. Typically, an inhomogeneous uptake of the radionuclide will be seen with relatively cold and hot areas in an enlarged thyroid gland, compatible with multinodular goitre (see Chapter 3.1.6). Ultrasonography accurately assesses the size of a goitre in the neck, but is of no use in substernal goitres. CT scans have diagnostic value for substernal goitres and characteristic features are: (1) anatomical continuity with the cervical thyroid, (2) focal calcifications, (3) precontrast attenuation of about 15 Hounsfield units greater than muscle due to the high iodine content of thyroid tissue, and (4) prolonged contrast enhancement after the administration of iodinated contrast material (6). Iodine-containing contrast agents in this setting, however, carry a risk of inducing thyrotoxicosis (see Chapters 3.2.4 and 3.3.1). MRI scans may provide similar valuable information on the extension of the goitre in relation to neighbouring structures. Chest radiographs may reveal an intrathoracic goitre by a smooth or nodular superior mediastinal paratracheal mass. Displacement and/or compression of the trachea is a frequent finding on radiographs, and tracheal compression is noted in 25–33% of patients (2).

Spirometric pulmonary function tests can be helpful. Visual inspection of the flow volume loop is a sensitive method for detecting upper airway obstruction (7) (Fig. 3.5.2.3). After reduction of goitre size, a significant increase in peak inspiratory and expiratory flow occurs. Inspiratory airflow is more severely affected than expiratory flow because of the tendency of the extrathoracic trachea to collapse on inspiration. Large goitres may contribute to obstructive sleep apnoea syndrome (8). Laryngoscopy can be performed to assess vocal cord mobility; rarely, acute angulation of the larynx is observed, making intubation hazardous. Barium oesophagograms and cine films are seldom useful in the diagnosis or management of compressive goitre.

Thyroid cancer is found in 4–17% of multinodular goitres depending on how carefully the surgical specimens are examined. Of 107 patients operated on for a benign multinodular goitre without suspicion of malignancy before surgery, 7.5% harboured incidental carcinomas, with papillary carcinoma being the most common variety (9). Substernal goitres also harbour malignant (mostly occult) cancer in 7% (4). It is doubtful, however, if these incidental cancers adversely affect life expectancy. Fear of malignancy is in general not warranted for women with a history of long-standing slowly growing multinodular goitre and family members with the same condition. However, in patients with a dominant cold nodule, a fast-growing nodule, or a nodule with a very firm texture, fine-needle aspiration cytology is indicated to exclude malignancy (10). Ultrasonographic characteristics of nodules may help in selecting nodules which should be biopsied.

Options in the management of patients with nontoxic multinodular goitre are simple observation, surgery, l-thyroxine, and radioactive iodine (11).

Data on the natural history of the disease indicates a gradual increase in goitre size by about 4.5% per year, under simultaneous development of increasing thyroid nodularity and thyroid autonomy (1). Long-term outcome studies report development of thyrotoxicosis (toxic multinodular goitre or Plummer’s disease) in 10% after a mean follow-up period of 4–5 years. (12, 13).

The big advantage of thyroidectomy is that it rapidly and effectively removes the goitre, albeit at the expense of a low but unavoidable morbidity. Surgical complications include postoperative haemorrhage in 0.5%, vocal cord paresis in 1–2%, and hypoparathyroidism in 2–4%, dependent upon surgical skill and experience. Persistent voice disabilities (dysphonia, hoarseness, fatigue, or reduction of voice range) are not uncommon (15%), and late hypothyroidism occurs in 5–8%.

The reported incidence of recurrent goitre after surgery varies widely, ranging from 4% to 20%, and even 40% after 30 years. In general, the recurrence rate increases with longer follow-up (14) (Fig. 3.5.2.4). The determinants of postoperative recurrences remain largely unknown. Most studies find no difference in postoperative plasma TSH and T₄ concentrations between recurrent goitre patients and those without, the presence of thyroglobulin antibodies and thyroid peroxidase antibodies, the type of surgery (unilateral or bilateral resection), the extent of lymphocytic infiltration in the surgical specimen, or thyroid remnant size (15–17), although a larger thyroid remnant size in the recurrent group is sometimes observed (18). One study indicates a higher frequency of a family history of thyroid disease in the recurrent goitre patients, implying a role of still unknown genetic factors (16). The relatively high recurrence rate has led some surgeons to advocate removal of all nodules found at intraoperative digital palpation, or even total thyroidectomy (19).

Whether or not postoperative treatment with thyroxine prevents recurrent goitre, remains controversial. Most open uncontrolled studies find that the recurrence rate is not lowered by T₄ medication (14, 16) (Fig. 3.5.2.4), and the same conclusion is reached in the few randomized but not placebo-controlled studies (17). Although these results are compatible with the finding that postoperative growth of the thyroid remnant is to a certain extent independent of TSH (20), serum TSH was not really suppressed in these studies. Italian studies (in iodine-deficient areas) observed a lower recurrence rate with TSH-suppressive doses of T₄ than with TSH-nonsuppressive doses (21), and additional benefit of adding iodine to T₄ treatment (22). In contrast, another large study found no preventive effect of T₄ despite suppressed TSH values (18). The available data do not support the routine use of T₄ in order to prevent postoperative recurrent goitre.

The rationale of T₄ treatment is to suppress TSH. TSH as a stimulus of thyroid growth is thought to play a permissive role in the pathogenesis of sporadic nontoxic goitre (see Chapter 3.5.1). In addition, administration of T₄ significantly inhibits growth of human multinodular goitre tissue transplanted to nude mice (23). The effect varies, however, in different specimens of nontoxic goitre, indicating a varying degree of autonomous replicating activity. Older observational studies report reduction of goitre size upon thyroid hormone medication in about two-thirds of cases; the response was better in diffuse than in nodular goitres (11).

In a placebo-controlled double-blind randomized clinical trial in patients with sporadic nontoxic multinodular goitre, the T₄ dose (initially 2.5 μg/kg per day) was aimed at TSH suppression and adjusted accordingly, resulting in a mean daily dose of 175 μg. A response was defined as a decrease of goitre size of more than 13% (the mean plus two standard deviations of the coefficient of variation of thyroid volume measurements by ultrasonography) (24). There were 5% responders in the placebo group and 58% responders in the T₄-treated group. After 9 months of treatment, goitre size had increased by 20% in the placebo group, remained almost the same in T₄ nonresponders, and decreased by 25% in T₄ responders; after discontinuation of T₄ treatment, the goitre grew again (Fig. 3.5.2.5). Thus, T₄ treatment in the so-called T₄ nonresponders arrested goitre growth, and continuous T₄ treatment is necessary to maintain its therapeutic effect. Goitre reduction by T₄ treatment was not related to pretreatment characteristics such as age, family history of thyroid disease, duration and size of the goitre, or radio-iodine uptake (24, 25). Only a pretreatment plasma TSH lower than 0.4 mU/l or insufficient TSH suppression during T₄ treatment seems to be related to a less favourable outcome. The reduction in the size of multinodular goitres is largely accounted for by a decrease in the combined nodular volumes (26).

The optimal degree of TSH suppression is not well established, but it seems reasonable to aim at TSH values of 0.1 mU/l. This means the induction of subclinical hyperthyroidism, which is poorly tolerated by some patients, requiring reduction of the T₄ dose in about 25% (24, 25). Subclinical hyperthyroidism carries a risk of atrial fibrillation and bone loss (see also Chapter 3.3.4).

There is renewed interest in ¹³¹I treatment of nontoxic multinodular goitre. The median ¹³¹I dose reported in the literature, is 1416 MBq or 4.6 MBq/g thyroid (125 µCi/g) corrected for 100% 24-h uptake (25, 27–30). The relatively low thyroidal radio-iodine uptake necessitates the use of high doses of ¹³¹I. An estimate of goitre size is required for dose calculation, preferably by ultrasonography (31). A large dose of ¹³¹I may require hospital admission for a few days; however, fractionation of the total radio-iodine dose over several months is feasible without jeopardizing outcome, allowing for treatment as an outpatient (32).

Iodine-131 therapy is very effective. The goitre (mean initial size of 126 g), shrinks in 94% of patients; the mean reduction in goitre size is 45%. The greatest fall in goitre size is observed in the first year after treatment; no further reduction is seen after 2 years (Fig. 3.5.2.6). Obstructive symptoms and signs are also favourably affected: in patients with large compressive goitres (including some with intrathoracic goitres), 1 year after ¹³¹I therapy the maximal tracheal deviation had decreased by 20%, the smallest cross-sectional area of the tracheal lumen had increased by 36%, dyspnoea and inspiratory stridor had improved in 8 of 12 patients, and compression of the superior vena cava had disappeared in 2 of 2 patients (28).

Independent variables determining the effect of ¹³¹I therapy are the administered ¹³¹I dose and initial goitre size; age and goitre duration are dependent variables, both being directly related to initial goitre size (30). The ¹³¹I dose required for a 50% reduction of goitre size is 4.8 MBq/g thyroid (29, 30). The larger the goitre, the lower the reduction in goitre size (30). Nonresponders and those with late recurrence of goitre growth (8% at 3–5 years after ¹³¹I therapy) have larger goitres and more often dominant nodules than responders (30). A second dose of ¹³¹I seems to be as beneficial as the first treatment (30).

An increase of obstructive symptoms after ¹³¹I treatment is often warned about but rarely seen, even in patients with large compressive goitres (28). Serial measurements of goitre size for 5 weeks after ¹³¹I therapy did not demonstrate a significant increase of thyroid volume, the maximum increase in the median volume being 4% on day 7 (33). Routine administration of prednisone as a preventive measure is thus not warranted.

Serum free T₃ and free T₄ indices increase transiently by 20% at day 7, reducing to 13% at day 14, and returning to baseline values at 3 weeks (33). Radiation thyroiditis with tenderness of the neck and slight thyrotoxic symptoms develops in the first few weeks after ¹³¹I treatment in 4% of patients. In view of its self-limiting and mostly mild nature, treatment is usually not necessary but salicylates (or, rarely, corticosteroids) can be applied successfully.

Graves’-like hyperthyroidism occurs in 4% of patients, usually developing 3–6 months after ¹³¹I therapy, and is related to the new appearance of TSH-receptor antibodies triggered by radiation-induced release of antigens from the thyroid (34, 35). The hyperthyroidism may be severe and may require treatment with antithyroid drugs for several months. The presence of thyroid peroxidase antibodies before treatment increases the risk of developing this complication (34).

The incidence of postradio-iodine hypothyroidism varies widely between studies. A cumulative 5-year risk of 22% is reported in one study (27), in good agreement with 25% hypothyroidism after 2–9.5 years reported in the literature. Determinants are the presence of thyroid peroxidase antibodies, a family history of thyroid disease, and a relatively small goitre (30, 34).

Large doses of ¹³¹I carry a theoretical risk of cancer development. A dose of 1.9 GBq (51 mCi) has a calculated 1.6% life-time risk of development of cancer outside the thyroid gland (36). When applied to people of 65 years and older the estimated risk is approximately 0.5% (36, 37).

Recent experimental studies have investigated whether the outcome of ¹³¹I therapy could be improved by prior administration of recombinant human TSH (rhTSH). The use of rhTSH overcomes the problems of a low radioactive iodine uptake (RAIU) (rhTSH significantly enhances the absorbed thyroid ¹³¹I dose) (38) and of an irregular RAIU (rhTSH gives a more homogeneous distribution of RAIU) (39). One could use rhTSH aiming at lowering the radiation dose of absorbed ¹³¹I; indeed, a single low dose of 0.01 or 0.03 mg rhTSH intramuscularly allows for a 50–60% reduction of the usual therapeutic ¹³¹I dose without compromising the effect on goitre reduction (39). Alternatively, one could use rhTSH to enhance the efficacy of ¹³¹I. Indeed, a number of placebo-controlled randomized clinical trials have demonstrated a larger reduction in goitre size after a single relatively high dose of 0.30 or 0.45 mg rhTSH intramuscularly (40–42). However, this is obtained at the expense of more adverse events, a higher rate of transient thyrotoxicosis in the first 3–4 weeks, and late hypothyroidism (40–43). Thus, the most appropriate schedule taking full advantage of rhTSH has not yet been established, and the results of a new formulation of modified-release rhTSH for use in nontoxic goitre is eagerly awaited.

Indications for treatment are listed in Box 3.5.2.2. The preferred treatment in a particular case requires much deliberation between patient and physician, taking into account the efficacy and side effects of each type of treatment (Table 3.5.2.1).

Surgery is the treatment of choice if malignant growth is suspected. It can also be considered the standard therapy in the case of large compressive or substernal goitres, but radio-iodine is a suitable alternative to surgery in elderly patients and those with cardiopulmonary disease (37). Radio-iodine is a perfect choice for patients who already have a suppressed TSH, a condition which precludes the use of thyroxine, but the efficacy of radio-iodine is smaller if a dominant nodule or a very large goitre is present.

In symptomatic patients who are younger and have smaller goitres, an alternative to surgery is thyroxine. The efficacy of thyroxine in reducing the size of multinodular goitre is low, and requires the continuous administration of TSH-suppressive doses raising concern about its long-term safety. Radio-iodine, although not devoid of side effects, has been offered as an alternative in view of its greater efficacy and better tolerance than thyroxine (25); it should probably be restricted to patients older than 40 years. Future developments may decrease the theoretical risk of radiation-induced cancer by the use of rhTSH in conjunction with radio-iodine, hopefully allowing a lower dose of ¹³¹I.

In the asymptomatic patient, a wait-and-see policy seems to be prudent. Although the natural history is characterized by continuous growth, large variation exists in the growth rate between individuals. If progressive goitre growth is observed during follow-up, intervention to prevent development of compressive and/or toxic goitres should be considered. In this respect, early rather than late intervention is advantageous when choosing radio-iodine, in view of the greater efficacy and lower radiation burden of ¹³¹I if the goitre is still relatively small.

The management of nontoxic multinodular goitre should be tailored according to the individual patient’s need, taking into account sex, age, symptoms, goitre size and texture, and serum TSH. However, no consensus exists on the most appropriate treatment for a particular patient, as was evident from a recent questionnaire among thyroidologists worldwide asking how they would treat a 42-year-old women with a multinodular nontoxic goitre of 50–80 g causing moderate local neck discomfort in the absence of clinical suspicion on malignancy (44). Of the respondents, 35% opted for no treatment at all, 42% for thyroxine, 15% for surgery, 5% for radio-iodine, and 3% for stable iodine. Marked differences in preferences existed between countries.