Surgery of Cervical and Substernal Goiter

Goiter Definition

First, it is important to come to an understanding regarding what a “goiter” is and to define “big.” This is not easy when one looks critically at the literature. Both greatest diameter and goiter weight have been used to define thyroid enlargement. In the studies, methods for determining goiter size range from physical examination measured in centimeters, to physical examination estimated in grams, to surgical specimen measured in centimeters or grams. Preoperative imaging diameters may also be used.

The definition of goiter varies substantially among reports. McHenry suggested 80 g and Russell 100 g, whereas Clark proposed 200 g as the threshold value. Studies investigating radioiodine treatment for multinodular goiter often define significant goiter as greater than 100 g. Hegedus, Nygaard, and Hansen found that goiter surgical specimens averaged 30 g for unilateral resection and 64 g for bilateral resection. In the study by Katlic, Grillo, and Wang, the average weight of substernal goiter was 104 g (range 25 to 357 g), with the greatest diameter averaging 9 cm (range 5 to 19 cm). In a series of more than 200 cervical and substernal goiters treated at Massachusetts Eye and Ear Infirmary and Massachusetts General Hospital, the mean weight was 143 g and the mean goiter size was 10.5 cm.

The largest goiter we can find documented in the literature is reported by Manoppo, who described a benign, nontoxic goiter that was 75 × 60 × 45 cm. The patient was unable to walk or sit because of the size of the goiter. Treatment was with right hemithyroidectomy under local anesthesia without major complications.

The World Health Organization (WHO) 1960 grading system for clinical assessment of goiter defines stage 0 as no enlargement; stages 1 to 3 describe progressive goiter enlargement. Stage 1A includes patients with palpable abnormalities; stage 1B includes patients with palpable and visual abnormalities with the neck in extension. Stage 2 is defined as a goiter that is visible with the neck in neutral position, and Stage 3 as a goiter that is able to be visualized at a considerable distance. The WHO 1994 goiter classification system is more streamlined. Grade 0 is defined as no palpable or visual abnormality. Grade 1 is defined as a palpable thyroid mass that is not visualized with the neck in neutral position and grade 2 as a visually apparent mass with the neck in neutral position.

Substernal Goiter Definition

Substernal goiter and its subtypes have been variously termed retrosternal, subclavicular, intrathoracic, mediastinal, aberrant, wandering, and spring goiter, as well as goiter mobile and goiter plongeant. Numerous definitions and classification schemes have been proposed for substernal goiter.

Lahey and Swinton defined substernal goiter as a “gland in which the greatest diameter of the intrathoracic component by x-ray was well below the upper aperture of the thoracic inlet.” In 1939, Crile simply defined substernal goiter as a lesion extending to the aortic arch. Lindskog and Goldenberg in 1957 defined substernal goiter as a goiter whose lower border radiographically reaches the transverse process of the fourth thoracic vertebra or lower. Katlic, Grillo, and Wang described substernal goiter as greater than 50% of the goiter present substernally. Sanders et al. defined substernal goiter as that which requires mediastinal exploration and dissection for removal. Other definitions have been offered.

Several workers have offered substernal classification schemes. Higgins based his classification scheme on the percentage of goiter in the neck versus the percentage of goiter in the chest, with less than 50% in the chest being described as incomplete intrathoracic, greater than 80% in the chest as complete intrathoracic, and other categories, including extension below the sternum (substernal [complete or incomplete]) and extension below the clavicle (subclavicular [complete or incomplete]). Cho, Cohen, and Som offered a grading system relating grade to percentage of goiter within the chest. Grade I is defined as 0% to 25% of the goiter within the chest, grade II as 26% to 50%, grade III as 51% to 75%, and grade IV as greater than 75%. Shahian offered an interesting and detailed classification scheme. In this classification scheme, type I substernal goiter is associated with anterior mediastinal extension; type IA involves “isolated” anterior mediastinal disease, whereas type IB involves “extensive” substernal involvement. Type II involves posterior mediastinal involvement, with type IIA being an isolated posterior mediastinal goiter, type IIB being a posterior mediastinal goiter with ipsilateral extension relative to the thyroid lobe of origin, and type IIC being a contralateral extension relative to the thyroid lobe of origin, with C1 being retrotracheal and C2 being retroesophageal.

A classification system for substernal goiters is most useful when it takes into account the features of substernal goiters that must be appreciated to extract these goiters safely. We define substernal goiter simply as a goiter that is associated with substernal extension such that the thoracic component requires mediastinal dissection to facilitate extraction. We believe that all substernal goiters require axial computed tomography (CT) scanning to differentiate between the various subtypes. Such differentiation provides tremendously useful surgical information. We propose the following substernal goiter classification scheme ( Table 6.1 ).

Prevalence

Multinodular goiter affects 4% of the United States’ population and up to 10% of the British population. New thyroid nodular disease occurs in 0.1% to 1.5% of the general population per year. Globally, iodine deficiency contributes to the vast majority of cases of multinodular goiter and was estimated to affect 1.5 billion people, or nearly 30% of the world’s population in 1990. Further, it is estimated that approximately 655 million people in 118 countries are affected by endemic goiter. Endemic goiter regions are defined as iodine-deficient regions in which at least 5% to 10% of the population is affected by goiter. In certain iodine-deficient regions, higher goiter rates occur. In 1994 in Bangladesh, approximately 47% of the population was affected by endemic goiter. The majority of the natural iodine supply exists as iodide in the world’s oceans. It is therefore mainly noncoastal mountainous and lowland regions—where iodine is leached from the soil by flooding, heavy rainfall, and deforestation—that are at risk for endemic goiter. Sporadic forms of multinodular goiter do occur in iodine-replete regions with lesser prevalence. Prevalence estimates of sporadic goiters vary between authors, ranging from less than 4% (clinical evaluation series) and between 16% and 67% (ultrasound series).

Based on tuberculosis screening radiography in Australia and the United States, substernal goiter has been estimated to be present in 0.02% of the general population and 0.05% of females older than 40 years. The incidence of substernal goiter was found to significantly increase with age, with 60% of substernal goiters occurring in patients older than age 60 years. Rates of substernal goiter in the past several decades appear to be decreasing, perhaps related to the introduction of iodized salt, thyroid hormone suppressive therapy, radioiodine use in selective cases, and perhaps more sensitive detection and earlier intervention. Substernal goiter, as a percentage of patients undergoing thyroidectomy, ranges from less than 1% to greater than 20%, depending on the series, with most suggesting a rate of approximately 10%. Substernal goiters represent approximately 5% of all mediastinal tumors.

Pathogenesis

The pathogenesis of goiter formation has classically relied on the notion that iodine deficiency promotes persistent elevated thyroid-stimulating hormone (TSH) levels, inducing diffuse thyroid enlargement through thyrocyte proliferation. Nodules form in the enlarged thyroid as the patient ages, eventually giving rise to multinodular goiter. Since the early 2000s, this classic model has been challenged by the view that the thyroid gland has an intrinsic propensity to form nodules over time. In this new model, low iodine levels and elevated TSH are considered additional factors, exacerbating the innate process of nodule formation.

The shift in conceptualization of goiter pathogenesis has stemmed from new cellular models, first proposed by Studer and Derwahl. These authors have argued that the development of multinodular goiter, at least in the later stages, is independent of TSH levels. Nodules arise because thyroid follicles are embryologically derived from polyclonal progenitors and thus have a heterogeneous sensitivity to TSH signaling. Similarly, thyroid follicles have a differential growth response to continuous exposure to goitrogens, explaining the multinodular pattern of goiters. In addition, thyrocytes may undergo somatic mutations and acquire a distinct growth capacity. There is some suggestion that oxidative stress (caused by iodine deficiency, smoking, or thyroid hormone synthesis) may lead to DNA damage, causing an increase in the spontaneous mutation rate and leading to tumorigenesis. The contribution of somatic mutations in multinodular goitrogenesis remains, however, somewhat controversial. This is in contrast with rare growth-promoting germline mutations of the TSH receptor gene, which are believed to be pivotal in congenital diffuse goiter with hyperthyroidism. Importantly, activating mutations in the TSH receptor gene have not been associated with an increased risk of malignancy.

Natural History

The natural history of untreated, sporadic, nontoxic goiter is not completely understood, but slow growth appears to be the general predictable pattern. Berghout et al. suggested a steady volume increase of up to 10% to 20% per year. Pregnancy, iodine deficiency, consumption of goitrogens, and alteration in suppressive or antithyroid medical regimens can result in goiter progression. Hemorrhage into a preexisting nodule can also result in the development of acute, regional, and airway symptoms. In patients presenting with diffuse goiter, there is a general tendency toward nodule formation and progressive autonomy, with hyperthyroidism ultimately developing in up to 10% of patients (see Chapter 8 , The Surgical Management of Hyperthyroidism).

Most substernal goiters arise in the setting of preexisting cervical goiter. It is of note, however, that some patients with substernal goiter have no significant cervical goiter component. Substernal goiters virtually always have a connection to the cervical orthotopic gland. Lahey, in his vast experience of approximately 24,000 goiter surgeries, believed that all substernal goiters arise from the cervical gland and maintain their cervical blood supply. Even in cases of extreme substernal goiter extending to the diaphragm, the mediastinal component has been found to contain connections to the cervical gland and a blood supply from the inferior thyroid artery. Connection to the cervical gland may be robust or attenuated, but it virtually always exists. The work of Torre, based on an impressive series of 237 substernal goiters, suggests that substernal goiters arise 10 years after cervical goiter presentation, suggesting that substernal goiter evolves from preexisting cervical goiter in most cases.

The inferior extension of cervical goiter and formation of substernal goiters are poorly understood. The inferior descent relates in part to the pattern of nodular disease within the cervical gland. Inferior progression results from a limitation of the strap muscles anteriorly, trachea medially, and vertebral column posteriorly. As Lahey and Swinton have described, the neck is “a space with no bottom.” The repetitive forces of deglutition, respiratory dynamics, negative intrathoracic pressure, and gravitational forces in the setting of permissive mediastinal and neck base fascial planes facilitate the downward extension of cervical goiter. Typically, anterior mediastinal extension (substernal goiter type I) occurs from the ipsilateral lobe’s inferior expansion. Descent associated with significant retrotracheal posterior mediastinal extension may arise from more posterior elements of the thyroid gland such as posterior tubercles of Zuckerkandl (see Table 6.1 ).

Cervical Goiter

Substernal Goiter

In many ways, the history and physical examination for patients with substernal goiter significantly overlaps with those for patients with cervical goiter. In our series we have found that as cervical goiter progresses substernally, given the restriction of the bony confines of the thoracic inlet, it increasingly compromises the airway. In short, substernal goiter evolution is strongly correlated with tracheal deviation, the development of regional airway symptoms, and radiographic airway compression. Buckley and Stark noted that although the maximum tracheal deviation with substernal goiter usually occurs at the thoracic inlet, it may occasionally occur farther inferiorly. Larger surgical series of substernal goiter show 70% to 80% of substernal goiter patients are symptomatic at presentation. Cervical mass is noted in 69% to 97%, respiratory symptoms in 42% to 96%, dysphagia in 26% to 60%, and acute airway presentation in 1% to 5% of patients. Interestingly, these series show that 10% to 30% of substernal goiter patients, as previously noted, have no significant palpable cervical abnormality, 3% to 7% present with vocal cord paralysis, and 4% to 50% are asymptomatic at presentation. Wax and Briant have noted that, with careful questioning, up to one-third of patients who are “asymptomatic” admit to symptoms.

Pemberton’s sign is described as the development of head and neck venous engorgement with facial congestion, plethora, and venous distention with arms raised over the head. It is sometimes expanded to include the development of transient respiratory insufficiency. Pemberton’s sign is thought to indicate goiter extension into the thoracic inlet, with secondary relative venous and airway obstruction. Our series of large cervical and substernal goiters suggests that Pemberton’s sign is insensitive in the evaluation of substernal goiter, because only 4.4% of patients presented with a positive Pemberton’s sign. Substernal goiters can also present with neck and upper chest pain and have rarely been associated with hematemesis secondary to downhill esophageal varices (without signs of portal hypertension), abscess formation, Horner’s syndrome, chylothorax (secondary to thoracic duct obstruction), transient ischemic attacks through “thyroid steal syndrome,” venous thrombosis, and intubation injuries, especially to the posterior tracheal membranous wall.

Laryngeal shift to the side of a dominant cervical goiter suggests contralateral substernal goiter and requires axial imaging of the neck and chest. Similarly, laryngeal shift without any palpable cervical findings suggests substernal goiter and similarly requires axial neck and chest imaging. Finally, substernal goiter is suspected when the clavicle intervenes before the inferior extent of the thyroid mass can be palpated.

Airway Assessment in Thyroid Disease

Foremost in goiter assessment is airway evaluation. The fundamental components of airway evaluation include determination of the rate and pattern of respiration, presence of sound with breathing (i.e., stridor), and voice quality. In patients with significant airway obstruction from thyroid disease, the initial assessment requires integration of a targeted but complete history and physical examination to expeditiously identify the site and magnitude of obstruction. One must always remember to keep in mind the overall global status of the patient with respect to his or her respiratory effort and signs of distress. Fatigue, restlessness, or apprehension can occur in the setting of hypoxia. The patient who is lethargic may be hypercapnic. Body position as an indicator of respiratory comfort, peripheral signs of oxygenation, and cyanosis should be carefully evaluated. Included in this assessment is the taking of vital signs and the use of pulse oximetry. One must be vigilant that a patient with a good pulse oximeter reading—or, for that matter, good arterial blood gas levels—may, a moment later, experience complete respiratory obstruction. The sound of respiration (i.e., the presence of stridor) gives an important clue as to the magnitude and location of airway obstruction. Stridor implies turbulent airflow through a stenotic airway segment. Significant extrathoracic obstruction (most commonly laryngeal, subglottic, or upper cervical tracheal) typically presents with inspiratory stridor and is seen as a variable extrathoracic obstruction on flow volume loop analysis with inspiratory phase flattening. Isolated expiratory stridor is seen in intrathoracic (lower tracheal) obstruction. Here, inspiration is silent and the voice is normal.

Acute Airway Compromise

The frequency of development of acute airway compromise in cervical and substernal goiter varies depending on the population studied. Unfortunately most of the information available is from surgical series. Allo and Thompson have estimated that from 1% to 3% of patients with untreated mediastinal goiter die of respiratory obstruction.

The airway can be affected by goiter through a number of mechanisms, including, most typically, ongoing slowly progressive airway deviation and compression. Acute events can also precipitate airway compromise, including hemorrhage into a nodule of multinodular goiter. Georgiadis has described acute stridor caused by hemorrhage within a nodule after neck trauma. Pulli and Coniglio have described a patient whose goiter increased in size after a fall on ice. Torres et al. have presented a number of cases showing acute life-threatening tracheal obstruction in patients with long-standing, intrathoracic goiter with subsequent histology showing multiple foci of recent hemorrhage. Hemorrhage as a mechanism of goiter size increase is supported by the work of Bodon and Piccoli. We have seen acute tracheal obstruction with magnetic resonance imaging (MRI) scan evidence of central nodular hemorrhage in what was ultimately found to be a large benign follicular adenoma ( Figure 6.2 ). Sudden airway symptoms may also arise as a result of cystic degeneration, malignant degeneration, upper respiratory tract infection, or pregnancy. Rare mechanisms for respiratory distress in patients with large goiters include decompensated right heart failure, pleural effusion, and pulmonary hypoperfusion secondary to compression of the pulmonary arteries. We agree with Cougard et al. that a patient with goiter and acute airway decompensation requiring endotracheal tube intubation should be brought to surgery when stable and while intubated. Ultimately, after thyroid surgery a laryngeal examination should be performed to rule out laryngeal edema and assess vocal cord function before extubation.

Miller et al., in a nonsurgical series of 400 patients with goiter, evaluated patients with pulmonary function tests and flow volume loops. Nearly one-third of such patients had flow volume evidence of upper airway obstruction. Gittoes et al., in a study of 153 goiter patients followed in a medical/endocrine clinic, also found that 33% of such patients had evidence of upper airway obstruction on flow volume loop analysis. Thus two large nonsurgical series suggest that up to one-third of patients with large goiters followed in endocrine clinics have evidence of upper airway obstruction as defined on flow volume loops. Limited data are available within this population as to the rate of development of acute airway symptoms. Alfonso found, in a surgical series of 91 patients with benign goiters with either radiographic or symptomatic evidence of upper aerodigestive tract compression, that approximately 9% of patients presented with acute upper airway obstruction. Reeve, Rubenstein, and Rundle, in a large screening radiographic study that identified patients with significant thyromegaly as defined by plain radiographs, found 7.6% of such patients presented with “profound respiratory obstruction.” In a surgical series of patients with substernal goiter, reports vary, with 1.3% to 5% of such patients presenting with acute airway insufficiency. Higher rates of acute airway insufficiency have been reported in other surgical series.

Thus approximately one-third of patients with goiter in medical series have upper airway obstruction as defined by flow volume loops, and from 1.3% to 9% of such patients may present with acute airway symptomatology. Unfortunately, a long, chronic, stable history does not preclude spontaneous acute airway insufficiency. Warren has beautifully documented cases of acute respiratory failure secondary to goiter in elderly patients. All presented with acute respiratory collapse without a history of respiratory symptoms 48 hours before the respiratory failure. Cho, Cohen, and Som also emphasized, in a review of 70 patients with substernal goiter, the potential for sudden and unpredictable respiratory distress.

There is sometimes a tendency to follow up patients with large compressive goiters with flow volume loops and symptomatic assessment. There is no evidence to suggest this is a rational approach. It is known that flow volume loops begin to detect airway obstruction when tracheal diameter is reduced to an extremely limiting 5 mm. Symptoms of acute airway insufficiency may not occur until up to 75% of the tracheal lumen is obstructed. At our institutions, serial axial CTs are recommended to determine surgical candidacy, most typically by documenting substernal extension or tracheal compression. We have found that regional signs and symptoms, such as upper airway obstruction or radiographic evidence of tracheal or esophageal compression, are more likely with masses larger than 5 cm. Also, fine-needle aspiration (FNA) represents a less accurate assessment with lesions of this size compared with thyroid nodules between 1 and 3 cm.