Changing concepts in the pathogenesis and management of thyroid carcinoma


  • Dr. Robert F. Gagel MD,

    1. Gagel is Professor and Chief in the Section of Endocrine Neoplasia and Hormonal Disorders at the University of Texas M. D. Anderson Cancer Center in Houston, Texas
    Search for more papers by this author
  • Dr. Helmuth Goepfert MD,

    1. Goepfert is Professor and Chairman in the Department of Head and Neck Surgery at the University of Texas M. D. Anderson Cancer Center in Houston, Texas
    Search for more papers by this author
  • Dr. David L. Callender MD

    1. Callender is Assistant Professor and Deputy Chairman in the Department of Head and Neck Surgery at the University of Texas M. D. Anderson Cancer Center in Houston, Texas
    Search for more papers by this author


Thyroid cancer is a relatively uncommon neoplasm, accounting for about 1.5 percent of all cancers in the United States. This year about 15,600 new cases will be diagnosed in the United States.1 Death related to thyroid cancer is also uncommon, with no more than about 1,200 deaths annually in the United States.1 The relative infrequency of death from this disease has led to a viewpoint that thyroid carcinoma is an innocuous tumor. However, this is an incorrect assessment. It is not always possible to determine which patients will develop aggressive and/or recurrent disease,2 and aggressive thyroid cancer is difficult to manage and associated with a high level of morbidity and mortality.

In the past five years, significant advances have been made in understanding the causes of thyroid carcinoma and improving methods for diagnosis and management. In this context, it is instructive to review the experience at a large referral center for thyroid carcinoma with the goal of understanding how technologic advances during the past decade have altered the concepts of cause and management of the several forms of thyroid carcinoma.3–11 This review will focus on several aspects of thyroid carcinoma, including classification and diagnosis, recent insights into the molecular pathophysiology of tumors derived from follicular and parafollicular epithelium, and a review of the experience with thyroid carcinoma at the University of Texas M. D. Anderson Cancer Center (UT-MDACC).

Classification, Staging, and Prognosis

Primary carcinomas of the thyroid gland are usually classified as differentiated thyroid cancer (papillary and follicular carci-nomas),12 medullary thyroid carcinomas, and undifferentiated or anaplastic carcinomas. Other less frequent classifications include Hürthle cell carcinomas, squamous cell carcinomas, lymphomas and other hematopoietic lesions, and a variety of unusual carcinomas and soft tissue sarcomas (Table 1).

Papillary thyroid carcinoma, the most common histologic type of thyroid cancer, accounts for about 60 percent of all thyroid cancers. More than 30 percent of patients with papillary thyroid carcinoma present with metastasis to regional lymph nodes, and about 10 percent of patients develop hematogenous metastasis.

Follicular carcinoma accounts for about 20 percent of all thyroid cancers, although lower frequencies have been reported recently.13 Hematogenous metastasis is more common for follicular carcinoma than for papillary carcinoma, and the prognosis is somewhat less favorable. Hürthle cell carcinoma, a malignancy derived from the follicular cell, has a prognosis similar to that of follicular carcinoma.4,14–21

The treatment of patients with differentiated thyroid carcinoma should be based on each patient's prognostic factors. Several classification and staging schemes have been introduced to facilitate identification of important prognostic variables that can guide the clinician. Some of the more widely used staging systems are summarized below.

The treatment of patients with differentiated thyroid carcinoma should be based on each patient's prognostic factors.


The simplest classification divides patients into low-risk and high-risk groups.15 The low-risk group includes patients without metastasis if they are male and younger than 41 years or female and younger than 51 years. Patients older than this without metastasis are considered low-risk if they are without extrathy-roidal papillary carcinoma, major tumor capsular invasion by follicular carcinoma, or a primary tumor larger than 5 cm in diameter. Patients who do not meet the criteria for low-risk disease are placed in the high-risk group.

Of 310 patients with differentiated thyroid carcinoma seen at the Lahey Clinic during a 20-year period, 89 percent were assigned to the low-risk and 11 percent to the high-risk group. Only 1.8 percent of patients in the low-risk group died during follow-up, while 46 percent of the high-risk patients died. The frequencies of recurrence were five percent for low-risk patients and 55 percent for high-risk patients. When the AMES classification system was applied to patients at UT-MDACC and the University of Chicago, 75 percent and 70 percent, respectively, were classified as low-risk.19,21 The 40-year survival probability for low-risk patients was about 95 percent compared with about 45 percent for high-risk patients.19


The International Union Against Cancer and the American Joint Committee on Cancer have adopted a tumor-node-metastasis (TNM) classification system. Age at diagnosis is also an important factor in this schema. Applied to the University of Chicago cohort, this system separates patients into four groups, each having a survival probability that differs significantly from the others.19 Eighty-two percent of patients were stage I with a 20-year survival of nearly 100 percent. On the other end of the spectrum, five percent of patients were stage IV with a five-year survival of only 25 percent. Similar findings were reported when 1,500 patients with papillary thyroid carcinoma were analyzed.22


The first version of the Mayo Clinic staging system for thyroid cancer incorporated a formula based on Age at diagnosis, histologic tumor Grade, Extent of disease at presentation, and tumor Size to calculate a prognostic score.23 The most significant prognostic variables, in descending order of importance, were distant metastasis, age at diagnosis, tumor size, extrathyroidal extension, and tumor grade. Patients in group I, who accounted for 85 percent of the total cohort, had a 20-year, disease-specific mortality of only one percent. The corresponding mortality was 20 percent for group II, 67 percent for group III, and 87 percent for group IV.22,23

Table Table 1. Classification of Thyroid Carcinoma
original image

Application of this system to other patient populations was difficult because of the infrequent use of tumor grading as done at the Mayo Clinic. Therefore, a second system was developed based on Metastasis, Age at diagnosis, Completeness of surgical resection, extrathyroidal Invasion, and Size.24 The MACIS score is calculated as 3.1 (for patients younger than 40 years) or 0.08 • age (if aged 40 years or older) + 0.3 • tumor size (in centimeters) + 1 (if incompletely resected) + 1 (if extrathyroidal extension) + 3 (if distant metastasis). Twenty-year mortality from thyroid cancer for group I (score less than 6) is one percent, 11 percent for group II (score 6 to 6.99), 44 percent for group III (score 7 to 7.99), and 76 percent for group IV (score 8 or more).


Another staging scheme for papillary thyroid carcinoma is the Clinical Class system proposed by DeGroot et al.25 Disease limited to the thyroid gland is assigned to class I, nodal disease to class II, extrathyroidal extension to class III, and distant metastasis to class IV. For 269 patients, disease-specific mortality was markedly increased in the two highest classes, but mortality for patients with class I and II disease was low, and the two groups did not differ.19


In the most recent update of Mazzaferri and Jhiang's series, 1,355 patients were analyzed retrospectively.26 Patients with stage 1 disease had tumors smaller than 1.5 cm; those with stage 2 disease had tumors 1.5 to 4.4 cm or cervical metastasis or more than three intrathyroidal foci; patients with stage 3 disease had tumors at least 4.5 cm in size or local tumor invasion; and patients with stage 4 disease had distant metastasis. Both 30-year recurrence and cancer mortality rates increased with higher stage. Variables present at diagnosis that predicted cancer-specific mortality for patients with stages 1 to 3 disease included increasing age, primary therapy delayed by at least 12 months, local tumor invasion, lymph node metastasis, tumor size, and male gender.

Any one multivariate prognostic scoring system may not be applicable to another patient population.27 Nonetheless, it is possible to draw general conclusions from the multiple staging systems that exist. Younger age at diagnosis, smaller size of primary tumor, absence of extrathyroidal extension, complete gross resection at the time of initial surgery, and lack of nodal or distant metastasis are factors common to most prognostic systems that portend low risk for tumor recurrence or disease-specific mortality. It is important to recognize that younger patients do occasionally have poor outcomes, and the generally good prognosis associated with this age factor must not be overly emphasized in patient management. The presence of extrathyroidal extension, distant metastasis, a more aggressive histologic subtype (e.g., the tall-cell variant of papillary thyroid carcinoma or oxyphilic follicular carcinoma), or surgically unresectable disease indicates a need for more aggressive therapy.

Factors That Increase the Risk of Thyroid Carcinoma


The major risk factor predisposing to papillary or follicular thyroid carcinoma is exposure to radiation. Many studies have documented the increased risk of thyroid carcinoma in individuals exposed to low-level radiation.10,28–31 The increase in the diagnosis of thyroid cancer between the 1930s and the 1970s is at least in part attributable to the widespread use of irradiation for treatment of a variety of head and neck disorders in the first half of this century.32

More recently the importance of radiation as a risk factor for thyroid carcinoma has been underscored by the startling increase in the diagnosis of pediatric thyroid carcinoma among children exposed to ionizing radiation following the Chernobyl nuclear disaster in the Ukraine in 1986. More than 100 cases of pediatric thyroid carcinoma were observed in the Gomel region of Belorussia between 1989 and the present in an area where no more than one to two pediatric thyroid carcinomas per year had been previously identified.33–35 These tumors were associated with local lymph node metastasis in about 75 percent of children and were poorly differentiated in more than 50 percent. There has been one death attributable to thyroid carcinoma in this group of children.36,37 What has been surprising about this experience is the rapidity of development of thyroid carcinoma in these children. One possible explanation is that the true radiation exposure in these children has been underestimated, a point for which there is some supportive evidence.38

Other factors that have been implicated in papillary or follicular thyroid carcinoma but are incompletely understood include the role of iodine deficiency39,40 and autoimmune thyroid disease.41,42

There is increasing evidence that genetic factors may play a role in a small percentage of papillary and follicular thyroid carcinomas. The well-known associations of Gardner's syndrome (familial colonic polyposis) and Cowden disease (familial goiter and skin hamartomas) with differentiated thyroid carcinoma provide well-defined examples.43–45 Papillary thyroid carcinoma may also occur with increased frequency in certain families with breast, ovarian, renal, or central nervous system malignancies, suggesting that insight into the causative genes for these disorders may lead to the identification of genes causative for papillary or follicular thyroid carcinoma.46–48


About 25 to 35 percent of all medullary thyroid carcinomas are identified as a component of one of the variants of multiple endocrine neoplasia type 2 (MEN 2). These clinical syndromes include multiple endocrine neoplasia type 2A (MEN 2A) (medullary thyroid carcinoma, pheochromocytoma, and hyperparathy-roidism),49,50 multiple endocrine neoplasia type 2B (MEN 2B) (medullary thyroid carcinoma, pheochromocytoma, mucosal neuromas, and marfanoid-like features),51 and familial medullary thyroid carcinoma (Table 1).52

Medullary thyroid carcinoma is inherited as an autosomal dominant feature of these syndromes, and over 90 percent of individuals who inherit the gene for MEN 2 will develop medullary thyroid carcinoma at some point during life. There is some evidence to suggest a higher than normal incidence of medullary thyroid carcinoma in association with Hashimoto's thyroiditis, although the mechanism of transformation is not un-derstood.53,54

Molecular Events Involved in the Pathogenesis of Thyroid Carcinoma

Rapid progress in the identification of cancer-causing genes over the past five years has led to a preliminary outline of molecular events likely to be important in the genesis of benign or malignant transformation of the follicular or parafollicular cells of the thyroid gland. Analogous to other neoplasms, the genes involved in the causation of thyroid carcinoma form a subset of important cell growth and differentiation regulatory factors that can arbitrarily be separated into membrane and nuclear factors (Table 2). The discussion below will focus primarily on the role of the RET proto-oncogene in the genesis of thyroid carcinoma because molecular analysis of this gene in medullary thyroid carcinoma has important clinical implications. A detailed discussion of the causative genes involved in differentiated55,56 or medullary thyroid carcinoma57 is provided by several recent reviews.


Perhaps the most notable example of the involvement of membrane-related signal transduction pathways in thyroid carcinoma is the role of the RET proto-oncogene in the genesis of malignant transformation both in follicular cells (papillary thyroid carcinoma) and parafollicular or C cells (medullary thyroid carcinoma). Two different mutational mechanisms have been implicated in the genesis of these tumors.

Table Table 2. Oncogenes Involved in the Pathogenesis of Thyroid Neoplasia
original image


The RET proto-oncogene encodes a tyrosine kinase receptor. This gene is not normally expressed in the thyroid follicular cell. As a result of one of several gross rearrangements,58–60 the tyrosine kinase portion of the RET proto-oncogene is brought under the control of a promoter for one of three genes expressed constitu-tively in the thyroid follicular cell (Fig. 1). The resulting chromosome 10 rearrangements have been given the name “papillary thyroid carcinoma oncogene” (RET/PTC1, 2, and 3).61 In each of these rearrangements, the normal regulatory sequences for RET and the sequences encoding the extracellular domains are lost (Fig. 1). This results in expression of the tyrosine kinase at a high level in the affected thyroid follicular cell.

Figure Fig. 1..

Molecular abnormalities of the RET proto-oncogene in papillary thyroid carcinoma (PTC) or hereditary medullary thyroid carcinoma (MTC). The RET/PTC oncogene is created by one of several chromosome 10 rearrangements that result in promoter sequences—D10S170 (H4), RIα subunit of protein kinase A, or ele1—driving the expression of the tyro-sine kinase portion of the RET proto-oncogene. In hereditary MTC, germline mutations of one of five cysteines in exons 10 or 11 (codons 609, 611, 618, 620, 634) have been identified in multiple endocrine neoplasia type 2A (MEN 2A) or familial medullary thyroid carcinoma (FMTC). A single germline mutation of the tyro-sine kinase region (exon 16) that converts a methionine to a threonine at codon 918 has been identified in 98 percent of patients with multiple endocrine neoplasia type 2B (MEN 2B). This same mutation occurs as a somatic (tumor only) mutation in about 25 percent of sporadic MTC.


The RET proto-oncogene is normally expressed in the thyroid parafollicular cell. Studies performed during the past three years have demonstrated germline point mutations of this gene in more than 95 percent of individuals with hereditary medullary thyroid carcinoma. These point mutations affect one of five cysteines (codons 609, 611, 618, 620, or 634) located in a cysteine-rich region of the RET tyrosine kinase receptor and activate it, thereby causing transfomation (Fig. 1).62,63 Although not yet proven, it is thought that mutation of one of these five cysteines causes activation of the tyrosine kinase receptor, thereby initiating the transformation event. Mutation of codon 634 is the most commonly observed mutation and is found in about 80 percent of all patients with hereditary medullary thyroid carcinoma.64–66

More recently families with familial medullary thyroid carcinoma have been found with codon 768 and 804 point mu-tations.67 A germline point mutation in the tyrosine kinase portion of the RET receptor (codon 918) has been identified in 95 percent of individuals with MEN 2B (Fig. 1).68 The codon 918 mutation has been identified as a somatic (tumor only) mutation in about 25 to 30 percent of individuals with sporadic medullary thyroid carcinoma.68

These discoveries have already had a great impact on the management of MEN 2. It is now possible to determine whether an individual in a family with a known MEN 2A or MEN 2B mutation is a gene carrier by straightforward DNA analysis.62,65,69 This makes it possible to exclude family members who are not gene carriers from further screening studies.

There are several lines of reasoning that have led workers in this field to suggest that total thyroidectomy should be performed around the age of six years in children who are MEN 2A gene carriers and shortly after birth in children with the MEN 2B mutation.64,66 The most compelling argument for thyroidectomy is the finding that the lifetime penetrance for medullary thyroid carcinoma in MEN 2A and MEN 2B are 90 percent and nearly 100 percent, respectively. A second reason for early thyroidectomy is that metastasis has been observed as early as age six years in MEN 2A and at birth in MEN 2B. Finally, experience from several groups, including our own, suggests that the risk of thyroidectomy in children aged six years differs little from that in adults.64,70

It is important to point out that the use of genetic information to manage hereditary medullary thyroid carcinoma is less than three years old, leaving open the possibility that management approaches may change as workers in the field gain greater experience. The approaches outlined above are a logical extension of early screening by pentagas-trin testing used during the last 20 years to identify and treat gene carriers at the earliest possible time point. A more comprehensive discussion of the issues is available.66


A number of other molecular abnormalities have been identified in differentiated and anaplastic thyroid carcinoma. None of these has assumed relevance for clinical management of thyroid carcinoma, but they are of likely importance in the pathogenesis of this neoplasm. Table 2 provides a list of these oncogenes and detailed references for the interested reader.71–87

Figure Fig. 2..

Lymph node regions of importance for management of thyroid carcinoma. (Reproduced with permission from Cancer of the Head and Neck, W.B. Saunders Company.)

Diagnosis of Thyroid Carcinoma

The identification of a thyroid nodule or mass is the most common presentation for differentiated thyroid carcinoma. Clinical features that raise the level of suspicion for thyroid carcinoma include new-onset hoarseness and vocal cord paralysis, hemoptysis, and extensive lymph node enlargement.

Examination of the neck is usually remarkable for a palpable nodule that is often clinically indistinguishable from a mass associated with a benign condition. Not infrequently in adults and especially in children, the initial manifestation of thyroid carcinoma may be a palpable lymph node in the neck. Palpable metastatic adenopathy is most often found along the middle and lower portions of the jugular vein (Fig. 2, regions II, III, and IV).11,88 Nodal disease is also commonly located lateral to the ster nocleidomastoid muscle in the lower portion of the posterior triangle overlying the scalene muscles (Fig. 2, regions IV and V).

Physical examination of a patient with a thyroid nodule should not be confined to the thyroid gland and the neck but should include the larynx, tongue, and cervical spine. Fiberoptic or indirect laryngoscopy should be performed to document vocal cord movement and to examine for the presence of ectopic thyroid tissue in the base of the tongue. Physical findings that might point toward a particular type of thyroid carcinoma include the presence of hypertension (medullary thyroid carcinoma), mucosal neuromas and marfanoid features (medullary thyroid carcinoma), and colonic polyposis (papillary thyroid carcinoma). Laboratory findings that point toward a particular diagnosis of thyroid carcinoma include hypercalcemia, hypercalciuria, and increased catecholamine production (medullary thyroid carcinoma).

A variety of diagnostic tests have been employed in an attempt to separate benign from malignant thyroid nodules, including radionuclide scanning, ultrasound, and fine-needle aspiration. Improvements in cytologic analysis over the past decade have made fine-needle aspiration the single most important procedure for assessment of a thyroid nodule.89–93

In a patient with a single thyroid nodule, the initial evaluation consists of thyroid function studies, including an ultrasensitive thyroid-stimulating hormone (TSH) measurement, thyroid antibodies, a serum calcium measurement, and a fine-needle aspiration of the palpable nodule. Ultrasound examination is performed when there is the clinical suspicion of multiple thyroid nodules, when the thyroid is difficult to evaluate by palpation, or to establish a baseline for following the size of the nodule.

A recent report suggests the usefulness of serum calcitonin measurements in the evaluation of thyroid nodules,94 although it seems clear that fine-needle aspiration is a more direct and cost-efficient method for diagnosis of medullary thyroid carcinoma. Thyroid scans, a mainstay of thyroid evaluation in the past, are now used infrequently to evaluate the thyroid gland because of their relative lack of discrimination between benign and malignant disease and the improved sensitivity of TSH assays, making it possible to detect an autonomously functioning thyroid nodule (hot nodule) or early hyperthyroidism by suppression of the serum TSH concentration. A thyroid scan is performed to identify a “hot” nodule in individuals with a suppressed serum TSH concentration.

The results of fine-needle aspiration provide the major determinant in the decision to proceed with surgery. Patients with a fine-needle aspirate indicative of malignancy are treated surgically. Individuals with a finding of a benign colloid nodule or thyroiditis are observed with or without thyroid hormone suppression. Further growth of the nodule while on thyroid hormone suppression is an indication for surgical removal. Surgical removal is also indicated when the fine-needle aspirate shows findings of a follicular neoplasm, because it is not possible to differentiate between benign and malignant follicular neoplasms with certainty without histologic examination of the entire nodule. The management of lymphoma or anaplastic thyroid carcinoma diagnosed by fine-needle aspiration is individualized and will not be discussed in this review.

It is important to emphasize that fine-needle aspiration is only a tool to be used by the clinician in the decision-making process. A decision to proceed with surgical exploration is made in up to five percent of cases where a benign fine-needle aspiration is obtained. Factors that may prompt a decision for surgical removal in the face of a benign-appearing fine-needle aspirate include the presence of a large goiter causing obstructive symptoms, the repetitive finding of a blood-filled cyst, a history of irradiation, or a family history of papillary thyroid carcinoma. In an occasional patient, anxiety regarding the possibility of thyroid carcinoma, uncalmed by a benign fine-needle aspiration result, may be an indication for surgical removal. Despite these occasional exceptions, there is clear evidence that the percentage of patients with thyroid nodules who receive surgical treatment has reduced over the past decade at our institution and others, resulting in a higher percentage of thyroid carcinoma diagnoses in the surgical procedures performed (Table 3).

Table Table 3. Specific Indications for Surgical Intervention for Thyroid Abnormalities
original image

We limit the use of computerized tomography (CT) or magnetic resonance imaging (MRI) to large or recurrent carcinomas suspected of invading surrounding soft tissue. These imaging techniques are essential for planning an operation in a patient with extrathyroidal extension of tumor and for determining the extent of lymph node metastasis. Neither CT nor MRI offers significant improvement of resolution over ultrasound examination, but bothprovide superior anatomic localization.95,96

Ultrasound examination of the thyroid and neck is an important technique for evaluation and long-term management of thyroid nodules. It is important, however, to understand the strengths and limitations of the technique to benefit most fully from its use. The technique provides the most sensitive method for characterization of thyroid nodules and lymph nodes.97,98 It is possible to determine the size, consistency (calcification or cyst), and number of thyroid nodules with certainty. In cases where a nodule or lymph node is difficult to palpate or there are multiple nodules, ultrasound-guided biopsy provides the greatest certainty for correct sampling of the nodule. The primary limitation of the technique is the necessity for a skilled operator, an individual often removed from the primary site where the patient is seen. Another limitation of ultrasound is its failure to provide anatomic guidance to the surgeon unless the surgeon participates in the imaging process.

Identification of a thyroid nodule discovered incidentally during ultrasound, CT, or MRI examination for another medical problem has occurred with increasing frequency over the past several years. In the milieu of a cancer center, these nodules most commonly are identified in a patient with an established primary malignancy of another organ previously treated with chemotherapy or radiation therapy. In addition to thyroid carcinoma, consideration must be given to the possibility of metastasis to the thyroid gland. In our institution a decision is made to proceed with fine-needle aspiration in most nodules greater than one cm in diameter.

The collective cost of the procedures described can be substantial. Sound management calls for deletion of procedures of marginal value in the evaluation of a thyroid nodule.

The Decision for Surgical Treatment

Despite the certainty that fine-needle aspiration brings to decision analysis, clinical judgment remains an important factor in the selection of patients for surgery. The presence of localized pain, dysphagia, or hoarseness suggests the possibility of malignancy. Rapid enlargement of a thyroid mass, particularly when associated with dyspnea, is indicative of a more aggressive local tumor growth and should prompt consideration for surgical resection.

A prior history of radiation exposure,10,28,30,31,99 age less than 20 years10,28,30,31,99 or more than 60 years,4,21,22 or growth of a thyroid nodule on suppressive therapy with thyroid hormone increase the likelihood of finding cancer in a thyroid nodule. The incidence of thyroid carcinoma in children or adolescents with a solitary nodule is as high as 40 percent, and there is some evidence to suggest that earlier intervention may improve prognosis. Other factors that should concern the clinician include the reaccumulation of fluid in a thyroid cyst, especially bloody fluid,100 or a dominant nodule in a patient with lymphocytic or Hashimoto's thyroiditis or Grave's disease.42,53,101 Specific indications for thyroid surgery are outlined in Table 3.

A total or near-total thyroidectomy should be performed for most patients with papillary, follicular, or medullary thyroid carcinoma.


There are occasional situations when a diagnosis of thyroid carcinoma is made during treatment of another condition. For example, the detection of a small focus of metastatic papillary thyroid carcinoma detected in a lymph node during dissection for squamous cell carcinoma of the head and neck generally requires no further therapy unless a thyroid tumor is detected. Our experience suggests that these tumors will remain clinically silent, particularly if patients are treated with suppressive doses of thyroid hormone. Similarly, the identification of a single, occult papillary thyroid carcinoma (<0.5 cm in diameter) during surgical removal of a portion of the thyroid for other reasons generally has little clinical significance.102,103 Our policy is to initiate thyroid hormone replacement in such patients and reevaluate these patients periodically.


A total or near-total thyroidectomy should be performed for most patients with papillary, follicular, or medullary thyroid carcinoma. A total thyroidectomy is defined as the removal of both lobes and isthmus with preservation of parathyroid glands and superior and recurrent laryngeal nerves. Exceptions to this recommendation include the finding of occult carcinoma or a papillary thyroid carcinoma less than 1.5 cm. We do not routinely perform a total thyroidectomy for multicentric carcinoma found on permanent sections after lobectomy for a less than 1.5 cm tumor.

Discussions regarding the necessity for total thyroidectomy are always controversial. In low-risk patients, a thyroid lobectomy may suffice for treating small (less than 1.5 cm and noninvasive) thyroid carcinomas. A recent study confirmed the safety of such an approach.104 We treat patients with a papillary or follicular thyroid carcinoma by total thyroidectomy when there is a history of previous radiation, gross disease in both lobes, or the presence of metastasis in regional lymph nodes or distant tissues. Our rationale for these recommendations is based on the prevention of local recurrence and the facilitation of postoperative treatment and long-term surveillance.21,105,106 Perhaps the most compelling argument for total thyroidectomy in papillary thyroid carcinoma is that central neck recurrence is less common after total thyroidectomy.16

Reasons for recurrence after lesser surgical procedures could include the presence of microscopic foci in the remaining thyroid lobe (found in 80 percent of thyroid glands in a whole organ section study performed at UT-MDACC over 30 years ago), which leads to recurrent cancer in the remaining lobe in 4.7 to 24 percent of cases. Although there is debate about the impact of recurrence on survival,22 40 to 50 percent of patients who die of thyroid carcinoma do so because of recurrent disease in the central compartment of the neck, and a high percentage of patients with recurrence in the thyroid bed (as high as 50 percent) will die of their carcinoma. Finally, radioactive iodine therapy for treatment of thyroid carcinoma is more effective in the absence of thyroid tissue, and complete removal of thyroid tissue makes it possible to use plasma thyroglobulin levels to screen for recurrent carcinoma during the follow-up period.

The rationale for total thyroidectomy for medullary thyroid carcinoma is based on several important facts. First, in a patient who presents with apparent sporadic medullary thyroid carcinoma, there is about a 10 to 15 percent chance it is hereditary and, therefore, bilateral and multicentric.107 Second, intrathyroidal metastasis is not uncommon in sporadic medullary thyroid carcinoma. In the patient treated by lobectomy with elevated postoperative calcitonin values, there is always the question of whether disease exists in the contralateral lobe (indicating either hereditary disease or intrathyroidal metastasis) or in extrathyroidal sites such as lymph nodes. These issues are more readily addressed if total thyroidectomy is performed at the time of primary surgery.107,108

An infrequent but recurring dilemma following thyroid lobectomy for an apparent benign nodule is the identification of papillary thyroid carcinoma in the nodule on the final histologic sections. Lobectomy is considered adequate treatment in this clinical situation if the papillary thyroid carcinoma is less than 1.5 cm in diameter with no evidence of multicentricity or metastasis. Our practice is to complete the thyroidectomy in all other patients.105


The primary surgical approach should focus on the thyroid lobe containing the suspicious nodule(s). During a meticulous dissection of the affected side, the recurrent laryngeal nerve is identified and protected, followed by resection of the isthmus and ipsilateral lobe. Parathyroid tissue is identified and preserved except in instances where there is extensive invasion by cancer or extensive metastasis in the paratracheal area. In patients with a nodule greater than 1.5 cm showing papillary thyroid carcinoma or evidence of local metastasis, a total thyroidectomy is completed by resection of the opposite lobe with particular care to identify and preserve parathyroid tissue and vasculature.

Figure Fig. 3..

Diagram illustrating the lymph node groups at highest risk for regional metastasis from differentiated thyroid carcinoma. (Reproduced with permission from Cancer of the Head and Neck, W.B. Saunders Company.)

We routinely examine the posterior surface of the thyroid gland for parathyroid tissue by loupe magnification and send a piece of any tissue identified for frozen-section examination. Tissue proven to be parathyroid gland is minced and implanted in a small pocket created in the sternocleidomastoid muscle.

We also advocate a compartmental lymph node dissection (interjugulo-paratracheal nodal dissection). Performance of this procedure during the primary surgical procedure makes it less likely to be required on a subsequent reexploration if there is a recurrence (Fig. 3).

It is important to preserve all parathyroid tissue, and if during the dissection the vascular supply of the parathyroid gland becomes compromised or is removed with the specimen, we recommend identification of it through frozen-section examination of a portion of it and reimplantation of the gland into muscle tissue, either at the surgical site or in the forearm.109

Judgment is required when balancing the necessity for a complete thyroid cancer removal and lymph node dissection against the possibility of permanent hypoparathyroidism caused by such a removal. The recurrent laryngeal nerves should be preserved whenever possible, and the superior laryngeal nerves, which usually run parallel to the superior vascular pedicle of the gland, should be identified and preserved as well.

Local invasion of tissues surrounding the thyroid gland is rare, but when present it is a significant cause of morbidity and mortality. It is important to define the extent of extrathyroidal extension of thyroid carcinoma and determine whether surgical removal is feasible. During primary surgery, it is important to determine whether it is necessary to resect laryngeal nerves, tracheal rings, or portions of the larynx.110–116 In most cases resection of these structures is indicated if there is involvement by thyroid carcinoma, and these indications are discussed in detail with the patient prior to surgery.117 Although radioactive iodine and external-beam radiation are excellent adjuvant forms of therapy, permitting narrower margins than are commonly employed in squamous carcinoma, the optimal goal of surgery is to completely remove all identifiable carcinoma.


A decision to proceed with neck dissection should be based on the type of thyroid carcinoma and evidence of tumor extension to local lymph nodes.11,17,96 There are several general observations that make decision making easier. Palpable papillary or medullary thyroid carcinoma metastasizes to regional lymph nodes frequently, while follicular thyroid carcinoma does so rarely. Careful preoperative assessment of lymph nodes by ultrasound and fine-needle biopsy or intraoperative inspection and biopsy is important to determine the necessity for dissection. Elective dissection of lymph node tissue for papillary thyroid carcinoma in the absence of demonstrable metastasis is of dubious value, whereas dissection is of value for cases in which metastasis is identified. In most cases it will be necessary to examine and biopsy central or other compartment lymph nodes to make a determination regarding the presence of metastasis (Fig. 3).

There are several lymph node groups that form the most likely route of lymphatic spread and should be considered for inclusion in a neck dissection.118 The so-called central or interjugular tracheal compartment forms the primary route for lymph node metastasis because of its proximity to the thyroid gland and pathway of lymph drainage. A second frequent route of lymphatic spread involves the nodes along the jugular vein from the subdigastric area to the root of the neck (levels II through IV, Fig. 2). A third common route of lymphatic spread is along the pathway of the inferior thyroid artery behind the common carotid artery and to the lower portion of the posterior triangle of the neck (level V, Fig. 2). A dissection strategy is planned to include these areas. This type of dissection differs from the classic radical neck dissection in that it does not include removal of internal jugular veins, sternocleidomastoid muscle, or the spinal accessory nerve. There is no evidence that a radical dissection improves outcome in most patients with thyroid cancer, although in a rare patient with direct extension of tumor into one of these structures, removal may be indicated. Dissection of the level I or submandibular triangle nodes is seldom necessary.

It is also prudent to perform a central compartment lymph node dissection in patients with palpable medullary thyroid carcinoma because of the high probability of regional lymph node metastasis.107,119 This is true for both hereditary and sporadic medullary thyroid carcinoma.120 Medullary thyroid carcinoma differs from most other head and neck tumors because the measurement of serum calcitonin provides a sensitive and relatively specific tumor marker.

A major question confronting the surgeon at the time of primary exploration is whether a more extensive lymph node dissection for medullary thyroid carcinoma will alter the clinical course of the disease. Experience in several centers over the past decade indicates that 15 to 20 percent of patients with limited nodal metastasis can be cured by extensive lymph node dissection, suggesting that consideration should be given to performing this procedure at the time of primary surgery.121 If extensive lymph node removal is considered, it is important to exclude the presence of distant metastasis prior to surgery by performance of CT, MRI, or octreotide scans of the neck, chest, and abdomen because extensive lymph node dissection is generally not indicated in patients with distant metastasis except for local control of disease.

Total thyroidectomy and lymph node dissection are generally well tolerated and may require a hospitalization of less than 24 hours for thyroidectomy and two to three days when combined with neck dissection. It is important to monitor the serum calcium concentration in the postoperative period and provide calcium and/or vitamin D supplementation if hypocalcemia develops. In an attempt to shorten the length of hospitalization, we begin calcium carbonate (1 to 2 g three times a day) and oral 1,25 dihydroxy vitamin D3 supplementation (0.5 μg orally three times a day) if the calcium remains below 7.5 mg more than 24 hours after the conclusion of the operation. If the serum calcium returns to normal in the first 48 to 72 hours postoperatively, this supplementation is discontinued. Thyroid hormone supplementation is generally deferred until a final decision has been made regarding adjunctive radioactive iodine therapy except in patients with medullary thyroid carcinoma where thyroid hormone replacement is begun in the immediate postoperative period.

Figure Fig. 4..

Impact of radioactive iodine (RAI) on recurrence and survival rates in papillary thyroid carcinoma. These results were updated and replotted from data presented in Samaan et al.21

Table Table 4. Schedule for Long-Term Management of Papillary Thyroid Carcinoma
original image

Postoperative Management of Thyroid Carcinoma


Experience over the past 40 years with radioactive iodine (RAI) therapy in many centers has demonstrated a significant effect on survival and recurrence rates in papillary and follicular thyroid carcinoma, although there is some controversy.22,122–124 The UT-MDACC recognized the potential value of RAI treatment and has used this therapeutic modality in more than 50 percent of patients treated for thyroid cancer over the past 35 to 40 years,4,21 although patients with thyroid carcinoma limited to the thyroid gland have been less likely to receive RAI.

The practice at the UT-MDACC for patients who have received a total thyroidectomy and have a tumor size greater than 2.0 cm has been to ablate residual thyroid tissue with 100 mCi RAI and to treat residual tumor with a dose of 150 mCi. Total-body radioactive iodine scans are repeated at six-month intervals until there is no uptake or the patient has received a total dose approximating 500 mCi RAI.

Figure 4 shows the effect of radioactive iodine therapy on both survival and recurrence in three groups of patients: those with disease limited to the thyroid gland, patients with metastasis to lymph nodes, and patients with extension to soft tissue of the neck. Although these survival curves are not adjusted for particular risk factors, in each of the groups radioactive iodine therapy had an effect on the rate of recurrence and survival. It is only in those patients with extension to soft tissue of the neck that the impact of radioactive iodine does not reach statistical significance.21 Although these data are retrospective and the populations included are heterogeneous, the results support a role for RAI in the postoperative management of papillary and follicular thyroid carcinoma and mirror results from other centers.

We routinely withhold thyroid hormone therapy for a four- to six-week period following total thyroidectomy in patients who have tumors greater than 1.5 to 2.0 cm and perform a total body scan following administration of 5 mci of sodium iodide I 131. Patients with positive scans are generally treated with 100 mCi of radioactive iodine or 150 mCi if there is obvious uptake in lymph nodes or an extrathyroidal area of metastasis. The patient is rescanned in six to eight months and retreated with 100 to 150 mci if there is residual (defined as greater than two percent of administered dose) or newly developed uptake. This process may be repeated several times, although a total dose of greater than 500 to 600 mci is rarely given. Most commonly one or two treatments with radioactive iodine will eliminate any residual uptake. The recent positive experience with recombinant TSH to stimulate radioactive iodine uptake in patients with thyroid cancer suggests that cessation of thyroid hormone may be unnecessary in the future.125,126


Papillary thyroid carcinoma is a rare cause of death, in part because of the generally benign course of the disease and in part because of the success of treatment of recurrence.26 Detection and treatment of such recurrences play an important role in the prevention of morbidity and mortality from papillary thyroid carcinoma.

General experience, including our own, suggests that no single diagnostic tool will detect all recurrences.97,125 We apply a series of overlapping strategies that include clinical examination, serum thyroglobulin measurements, ultrasound examination, and chest roentgenography on a periodic basis to detect recurrence of tumor. Ultrasound use has led to periodic identification of recurrent disease in patients with no evidence of RAI uptake and normal thyroglobulin levels. The frequency of follow-up examinations is greatest during the first three to four years following surgery, the time period during which a recurrence is most likely to occur, and less in subsequent years and decades (Table 4). The appearance of tumor recurrence in an occasional patient decades after the initial treatment suggests that surveillance should not be discontinued but performed less frequently when a decade or more has passed with no evidence of relapse.

We do not routinely perform total-body radioactive iodine scans for follow-up once the patient has a negative scan. Our experience indicates that most recurrent tumors following radioactive iodine therapy do not concentrate iodine, making regular RAI scanning of dubious value. When this observation is combined with the debilitating symptoms of hypothyroidism caused by discontinuance of thyroid hormone, enthusiasm for routine follow-up scans falls further. We do perform RAI scanning in patients with a rising thyroglobulin or other evidence of recurrent thyroid carcinoma because of the small possibility that radioactive iodine uptake may provide an additional therapeutic modality.

Long-term follow-up for medullary thyroid carcinoma is dependent on the extent of disease at the time of primary surgery. In those patients with intrathyroidal disease and no detectable calcitonin after a provocative pentagastrin injection, a periodic pentagastrin test measurement and clinical examination may suffice. In patients with local lymph node metastasis at the time of primary thyroidectomy and elevated calcitonin values postoperatively, we perform periodic ultrasound examinations to identify local recurrence.

Another important component of long-term management of medullary thyroid carcinoma is to identify individuals with hereditary medullary thyroid carcinoma. Preliminary analyses suggest that five to six percent of patients previously considered to have sporadic medullary thyroid carcinoma have RET proto-oncogene molecular abnormalities indicative of hereditary medullary thyroid carcinoma (R.F. Gagel, unpublished observations). Molecular analysis of the RET proto-oncogene, a commercially available test performed on a single blood sample from the affected patient, makes it possible to exclude hereditary medullary thyroid carcinoma with 99 percent certainty.64,66 It seems likely that this analysis will become a routine part of the evaluation of a patient with apparent sporadic medullary thyroid carcinoma. In individuals with a RET proto-oncogene mutation, the analysis should be expanded to determine whether the gene has been transmitted within the family. It is particularly important to study children and young adults because thyroidectomy in gene carriers at this age is most likely to result in cure.


The most common site of recurrence for papillary thyroid carcinoma is in lymph nodes of the neck. Recurrence is most commonly detected by clinical or ultrasound examination. The primary therapy is surgical removal of the affected node with consideration of a more extensive lymph node dissection on the side of recurrence if not previously performed. Four to six weeks following surgical removal of the recurrence and cessation of thyroid hormone therapy, a total-body radioactive iodine scan is performed to determine whether additional radioactive iodine therapy might be beneficial.

Management of recurrent disease in which there is invasion of soft tissue, larynx, trachea, esophagus, or other structures in the neck or upper mediastinum must be individualized. A wide variety of techniques have been developed to remove and reconstruct or replace parts of the trachea, larynx, or esophagus and are routinely employed at our institution.110–116 A discussion of the indication and use of these techniques is beyond the scope of this review. In patients with extension of tumor into neck structures, external-beam radiation is considered in addition to radioactive iodine therapy.21,127

Chemotherapy is considered in patients with a large papillary thyroid carcinoma following primary thyroidectomy, aggressive tumor that cannot be removed by surgical excision, or metastatic tumor that is disseminated and does not take up radioactive iodine.5,128


A wealth of knowledge regarding the molecular causation of thyroid carcinoma has been accumulated over the past five years. This information has already had a significant impact on the management of some forms of thyroid carcinoma. The challenge during the next 10 years will be to incorporate newly acquired information into diagnostic and therapeutic approaches to thyroid carcinoma and to coordinate use of this information with time-tested approaches to further decrease morbidity and mortality from thyroid carcinoma.