Full Example Grave's

Case Presentation B.H. is a 46 yo BF with a history of enlargement of the thyroid dating to 1978 when she initially presented to Reynolds Health Center for evaluation. At the time she was noted to have a bilaterally enlarged gland and was referred for thyroid scan after thyroid function studies were found to be normal. Thyroid scan revealed an asymmetrically enlarged gland with an area of decreased uptake in the lateral portion of the right lobe of the thyroid. No definitive therapy was instituted at the time. In 1980, she again presented for evaluation of the thyroid gland. She underwent ultrasound of the thyroid which revealed an asymmetric gland with a 1.5 cm solid nodule in the lateral portion of the right lobe. Thyroid function studies were again normal, and she was started on Synthroid. In 1982, she was evaluated via FNAB of the thyroid which revealed benign cytodiagnostic results. She was continued on Synthroid, but for the next 12 years her care was essentially provided through the walk-in clinic, where her Synthroid was intermittently stopped due to a "lack of response," only to be restarted by the next physician seeing her. When she presented in 06/95, she had not been taking Synthroid for five years. She presented to the walk-in clinic with complaints of her neck enlarging over the previous six to eight months. Clinical examination revealed an enlarged thyroid (two to three times normal) with a large nodule on the right. No adenopathy was noted; no physical exam findings of hyper- or hypothyroidism were noted. TSH was 0.7 mIU/L. She was placed on yet another attempt of thyroid hormone suppressive therapy with Synthroid 0.05 mg/d, and follow up appointment was arranged. She presented for routine history and physical examination in 08/95. She denied a history of head or neck irradiation, family history of thyroid malignancy, symptoms of hyper- or hypothyroidism, dyspnea, hoarseness, dysphagia, or neck pain. She admitted to a one year history of neck enlargement described as an inability "to button the top button on my blouse." She denied however, any recent significant rapid change in the size of the gland. PMH : postmenopausal, peptic ulcer disease Medications : Synthroid 0.05 mg/d Allergies : NKDA Premarin 0.625 mg/d Provera 2.5 mg/d Social History : remote history of tobacco use, none X several years

denies alcohol or drug use Family History : mother with benign nodular goiter, otherwise noncontributory PE : afebrile, wt 202 lbs, BP 120/84, HR 76 regular HEENT: no proptosis or lid lag noted; neck revealed an enlarged gland 2-3X normal, asymmetric with R>L lobes, isthmus enlarged. Surface was irregular with a 2 cm palpable nodule of the R lobe laterally, the lower poles were not palpable, no thyroid bruit, no tenderness noted. LN SURVEY : No palpable anterior cervical, posterior cervical, or supraclavicular adenopathy CHEST : clear to auscultation bilaterally HEART: regular rate and rhythm, no murmurs noted ABDOMEN : unremarkable EXTREMITIES : unremarkable for signs of hypo- or hyperthyroidism LABORATORIES : TSH 0.5 mIU/L, SMAC and CBC unremarkable

Her Synthroid dose was increased to 0.088 mg/d, and she was scheduled for follow up in 3 months. She returned for follow up in 12/95 without significant change in her symptomatology or clinical examination. She did admit to an increased feeling of self-consciousness over the past several months regarding her thyroid size. She denied symptoms of hyperthyroidism. At this time, the possible further management options were discussed with the patient, particularly with the need to evaluate the gland for possible malignancy. FNAB was suggested, as well as the other options of observation on or off Synthroid, as well as surgical excision and

its complications. She requested referral to a thyroid surgeon to discuss thyroidectomy due to her increasing self consciousness relating to cosmesis. She deferred repeat FNAB until after discussion with a surgeon. The patient subsequently underwent subtotal thyroidectomy in 1/96. Surgical pathology revealed a multinodular goiter with a dominant 2.7 cm nodule of the right lobe of the thyroid. There was no malignancy documented. She was placed on Synthroid postoperatively, but has not returned for follow up in the medicine clinic.

THE CLINICAL PROBLEM Thyroid nodules are a common clinical problem. Physicians must be aware of the need for further evaluation of thyroid nodules in order to rule out malignacy, even though the likelihood of finding cancer is low. A safe, expedient, and low cost evaluation is therefore of extreme importance4. In an effort to help the clinician, extensive literature has been written with regards to the evaluation and management of nodular thyroid disease. The overwhelming volume of literature, accompanied by the conflicting opinions of well respected authors, can easily confuse even the most astute clinician15. The hope here, is to provide a basic understanding of the clinical evaluation and management of nodular thyroid disease, with respect to efficiency, cost containment, and to avoid invasive procedures in those patients with benign nodular disease for whom risk of malignancy is low.

DIFFERENTIAL DIAGNOSIS OF thyroid nodules The differential diagnosis of the thyroid nodule is rather broad. A partial listing of the etiologies is presented in Figure 1.

differential diagnosis of the thyroid nodule Colloid (adenomatous) nodule Thyroid adenoma Follicular adenoma Hurthle Cell adenoma Thyroid Carcinoma

Thyroid Cyst Pure cyst Complex cyst Thyroiditis Acute

Primary

Subacute

Papillary

Hashimoto’s

Follicular/Hurthle Cell

Graves’ Disease

Medullary

Infectious diseases

Anaplastic

Abscess

Metastatic/Direct invasion

Tuberculosis

Renal Cell

Infiltrative/Granulomatous disease

Breast

Sarcoidosis

Lung

Amyloidosis

Melanoma

Developmental Abnormalities

Colon cancer

Thyroid hemiagenesis

Gastric carcinoma

Thyroglossal duct cyst

Head and Neck tumors

Teratoma

Hodgkin’s disease

Miscellaneous

Thyroid Lymphoma Figure 1. Adapted from Burch, in Endocrinology and Metabolism Clinics of North America, 19954.

Prevalence of Thyroid Nodules Thyroid nodularity is extremely common. The prevalence of thyroid nodularity has been estimated in numerous studies. Most authors agree that the prevalence by palpation lies somewhere in the range of 4-7% of the general population.9,17,19,22 Most thyroid nodules are not clinically recognized, as necropsy studies reveal that the prevalence of nodularity approaches 3757% of the general population4,6,19,20. Ultrasonography studies estimate the prevalence of thyroid nodularity at approximately ten times higher than that appreciated by palpation alone--i.e. 4050%.4,6 Also, patients felt to have solitary nodules by palpation were shown to have multiple nodules by ultrasonography in 16-48% of cases.4 The annual incidence of new palpable nodules is estimated at 0.1%. This rate increases approximately 20 fold in the event of prior radiation exposure.22 At this rate of development, approximately 250,000 new nodules would come to clinical attention each year.7 The prevalence of thyroid nodules is also dependent upon age and sex. Nodules are more frequently seen in females with prevalence ratios ranging form 1.2:1 to 4:1. A female:male ratio is documented at 2:1 in many studies. The prevalence of nodularity increases with advancing age--50% of men and women in their 6th decade, and greater than 80% of women and 65% of men in their 9th decades at autopsy were shown to have nodular thyroid disease.4 In contrast to the high prevalence of nodular thyroid disease, the presence of thyroid carcinoma is rare, accounting for approximately 5% of all palpable thyroid nodules.22 In the United States, thyroid carcinoma represents about 1% of all malignancies, accounting for approximately 12,000

new cancers each year,13,17 resulting in about 1000 deaths per year.17 Thyroid carcinoma occurs more frequently in females due to the greater prevalence of nodular thyroid disease. However, nodular thyroid disease in males is twice more likely to be malignant than in females.1 The likelihood of thyroid carcinoma is greatest in patients under 25-30 years of age and in those over age 60.4 Those individuals in the older age groups are also more likely to demonstrate aggressive (anaplastic) subtypes of thyroid carcinoma and have more extensive disease, and therefore a higher likelihood of dying than their younger counterparts.4 The lowest proportion of malignant nodules (2.9%) was observed in patients 31-40 years of age, and was twice as frequent (6.5%) in patients younger than 20-25 years of age. The prevalence of malignant nodules was 6 fold greater (16.4%) in patients over age 70.1 Until recently, it was thought that multinodular glands demonstrated lower rates of malignancy than glands with solitary thyroid nodules. Belfiore, however, demonstrated that at operation, 4.7% of 4500 patients with solitary nodules harbor malignancy. A similar proportion (4.1%) of 1100 patients with multinodular goiters harbored malignancy, thus dispelling the belief that multinodular goiters were at lower risk of malignant disease.2

Risk Factors for Thyroid Malignancy RISK FACTORS FAVORING THE MALIGNANT NATURE OF THYROID NODULE Age < 20 or > 60

fixation of the nodule to surrounding tissue

male sex

firm, hard, and irregular surface

history of thyroid cancer

lymph node enlargement

histroy of irradiation

hoarseness

presence of solitary nodule

dysphagia

rapid growth

dyspnea

Figure 2. Adapted from Gupta, in Clinics in Geriatric Medicine, 199510.

In addition to the risk factors of age and sex noted previously, other risks for malignancy include a history of exposure to ionizing radiation to the head, neck, and chest area, particularly as a child. Here, new nodules develop at a rate of about 2% annually (20 fold greater than in nonirradiated patients) after exposure to as little as 200 to 500 rads of ionizing radiation. The peak incidence occurs 15-25 years after exposure, but may be seen as early as 3-5 years following exposure.4 Both benign and malignant nodules are more frequently seen following exposure to radiation to the head, neck, or chest,13 and may be present in as many as one third of

all nodules in patients with prior history of irradiation.19 Irradiation to the head, neck, and chest was used in the period from the 1920’s to the late 1950’s to treat many benign conditions including tonsillitis, acne, hemangiomas, thymic enlargement, tinea capitis, impetigo, pertussis, keloids, Bell’s palsy, and lymph node hypertrophy.21 It has been estimated that as many as one million Americans have received radiation therapy to the head, neck, and chest, during this time period.4 There has been an increased incidence and prevalence of thyroid disease observed in residents of Hiroshima and Nagasaki following detonation of the atomic bomb in 1945. There has also been an acute increase in the incidence of thyroid nodules in the population near Chernobyl following the nuclear reactor breakdown in 1986.4 These nodules are of both the benign and malignant subtypes. Other risk factors for thyroid malignancy include historical and physical exam findings. A positive family history of benign nodular thyroid disease may be reassuring. However, a history of medullary thyroid carcinoma or MEN II syndromes, or papillary carcinoma increases the likelihood of a malignant finding.10,13,17 Local symptoms suggestive of compression, obstruction, or invasion including dyspnea, dysphagia, or hoarseness indicating recurrent laryngeal nerve involvement, are suggestive of malignancy.10,19,21 Patients with a rapid growth of a solid thyroid mass,19,20 anterior or posterior cervical or supraclavicular lymphadenopathy, fixation of the mass to surrounding tissues or structures, a firm, hard, irregular surface, and the presence of a solitary nodule in an otherwise normal gland or a dominant nodule in a multinodular goiter are more likely to be associated with malignancy.5,10,13,19,20 These findings lack sensitivity and specificity for thyroid malignancy.4 Lesions greater than 3 cm in diameter are also more likely to harbor undetected malignancy than smaller lesions. Hashimoto’s thyroiditis has been documented to be a frequent preexisting condition in patients who ultimately develop thyroid lymphoma.4 Patients with highly suspicious lesions (defined as a positive family history of medullary carcinoma, rapid growth, firm consistency, hoarseness, fixation to adjacent local structures, or enlargement of regional lymph nodes) had carcinoma in 71% of cases. Patients with two or more of these characteristics demonstrated carcinoma 100% of the time.4

Evaluation of Thyroid Nodules Laboratory Data Tests of thyroid function are useful only to the extent that they assist in defining the context in which nodular thyroid disease occurs. Thyroid function testing (serum TSH, thyroxine, and triiodothyronine) usually reveals normal results.9,17 Most authors would accept simply a serum TSH as an adequate screen of thyroid function. However, no laboratory test can reliably distinguish benign from malignant thyroid disease.20 A subnormal TSH may indicate a hyperfunctioning nodule, which should be evaluated further to rule out an autonomously functioning nodule. Conversely, an elevated serum TSH may indicate

hypothyroidism, which in the United States, in the absence of prior therapy for hyperthyroidism, is most commonly due to Hashimoto’s thyroiditis.19 The nodules may represent thyroid tissue stimulated to enlarge under the influence of high TSH, or they may represent true neoplasia. Antimicrosomal antibodies may be useful in the diagnosis of Hashimoto’s thyroiditis, if present in high titer.4 Serum thyroglobulin levels are not helpful in the differentiation of benign from malignant thyroid nodules,4 but may be useful in post surgical follow up evaluation of patients with thyroid carcinoma. Serum calcitonin levels are generally elevated in patients with medullary thyroid carcinoma presenting as a palpable mass, but may require stimulation with pentagastrin and calcium to detect disease in the earlier stage of C-cell hyperplasia.4 Although some authors have advocated the screening of all patients with palpable thyroid masses with calcitonin levels, this is not cost effective, since only 2-5% of all thyroid malignancies represent medullary thyroid carcinoma.9

Nuclear Medicine Scans Radioiodine isotopes, which are trapped and organified to tyrosine residues of thyroglobulin and subsequently stored in the colloid, are sometimes employed in the evaluation of thyroid nodularity. 131I was commonly used in the past and remains useful particularly in the evaluation of metastatic thyroid carcinoma. 125I has a lower radiation energy delivered to the thyroid than 131 I, but has a longer half life, and is rarely used. 123I has a short half life of approximately 13 hours, can be administered orally, lacks beta emissions, and produces a radiation dose to the thyroid of about 1% of that delivered by 131I. Radioiodine scans overall produce a greater clarity and lower background than technetium scans. Radioiodine isotopes are considerably more expensive, require a cyclotron for production, and take longer to perform. Imaging at 4 hours allows the procedure to be performed in a single day, but traditionally a 24 uptake scan is performed, making the procedure 2 days in duration.19 THYROID SCINTIGRAPHY ADVANTAGES

DISADVANTAGES

•

Will diagnose hot nodules, eliminating biopsy or surgery

•

Warm or cold nodules require further study

•

•

Cysts and hemorrhages appear cold

Will pick up aberrant thyroid tissue

•

•

Cost to perform

Able to diagnose multinodular goiter

•

•

Potential for radiation exposure

Will confirm substernal thyroid

•

•

May require two days to perform

Will identify cold nodule in a patient with Graves’ disease

Figure 3. Adapted from Dwarakanathan, in Comprehensive Therapy, 19935.

99m Technetium pertechnetate is a monovalent anion which is actively concentrated by the thyroid gland, but unlike iodine, undergoes negligible organic binding. Its half life is approximately 6 hours, making the radiation delivered to the thyroid very low. It is administered as an intravenous bolus, with imaging performed 30 minutes later. Many investigators feel that the convenience, availability, and reduced cost of 99m technetium pertechnetate make it an acceptable imaging agent, since the slightly better quality of radioiodine scanning is of little clinical significance.19 The major objective of radionuclide thyroid scanning with 131 I, 123I or 99m technetium pertechnetate is to classify nodules as hot, cold, or warm depending on their ability to differentially concentrate radioisotope. Most benign and virtually all malignant neoplastic or nodular thyroid tissue concentrate both 99m technetium pertechnetate and radioiodine less avidly than adjacent normal thyroid tissue, resulting in a cold appearance on thyroid scanning. Autonomously functioning tissue will frequently concentrate both agents more avidly than normal adjacent tissue, producing a hot appearance on scan. Three to 8% of benign or malignant nodules will concentrate 99m technetium pertechnetate, but fail to organify radioiodine. Such nodules appear hot, or warm / indeterminate on 99m technetium pertechnetate scans and cold on radioiodine scans. Most authors therefore suggest that functional nodules on 99m technetium pertechnetate scans be rescanned with radioiodine. 19 The role of thyroid scintigraphy in the initial evaluation of the thyroid nodule has recently been questioned by several investigators. Prior to the advent of FNAB, thyroid scintiscanning was of premier importance in guiding the management of nodular thyroid disease. Today, with the current utility of FNAB, the role of radionuclide thyroid screening has become equivocal. Radionuclide imaging studies cannot reliably distinguish malignant from benign nodules as demonstrated by the data presented by Ashcraft and Van Herle. They found that in a series of patients with nodular thyroid disease who underwent surgery and histologic examination regardless of the findings of radionuclide screening, 84% of nodules were cold, 10.5% were warm and 5.5% were hot -- at histologic exam 16% of cold nodules, 9% of warm nodules and 4% of hot nodules were shown to be malignant. 20 Nuclear medicine scanning of the thyroid gland has limited utility beyond the classification of nodules according to their ability to trap radioisotope. Because the overwhelming majority of benign and malignant nodules appear hypofunctional relative to adjacent normal thyroid tissue, the finding of a cold nodule has relatively low specificity. 4 Radionuclide scanning has relatively poor ability to differentiate malignant from benign lesions in nodular thyroid disease. Nuclear scans also have poor ability to define the status of nodules located in the periphery or isthmus of the gland17 due to the dependence of the technique on comparison of the area in question to adjacent thyroid tissue. Other problems associated with the use of nuclear thyroid imaging relate to the interpretation of a nodule as warm or indeterminate. The majority of thyroid cancers under 2 cm, as well as a minority of larger tumors, will fail to appear cold on scan because they are located anterior or posterior to normally functioning thyroid tissue. Conversely, autonomous nodules which are unable to adequately suppress pituitary TSH production will appear warm due to the ability of adjacent tissue to take up radioisotope. 17

Thyroid scintigraphy does, however, have utility in the evaluation of a clinically suspected autonomous nodule. Based on the assumption that normal tissue requires TSH to stimulate the uptake of radioisotope, autonomous nodules are able to concentrate the isotope in the absence of TSH. Although hot nodules rarely harbor malignancy, their propensity for subclinical or overt hyperthyroidism usually dictates further therapy with surgery or radioablation.4 Radionuclide scanning may also be useful in the evaluation of lesions found to be indeterminate on FNAB, as hyperfunctional nodules are rarely malignant.13 Indeterminate or suspicious lesions are most appropriately managed with surgical excision, as will be demonstrated in a later section of this paper. Patients with Graves’ disease and a dominant nodule on palpation that proves to be cold on scintiscanning, have a higher risk of malignancy and should be referred to surgery for near total thyroidectomy as definitive therapy.4

THYROID HORMONE SUPPRESSION THERAPY AS A DIAGNOSTIC TOOL Thyroid hormone administration may also be used as a diagnostic tool. This method rests on the assumption that benign thyroid nodules, but not malignant ones, depend on TSH for development and growth, and that TSH secretion is regulated by feedback inhibition by thyroid hormone. If thyroid hormone administered in suppressive doses produces regression or disappearance of the nodule, one would assume that the nodule is benign. If there is no change, or the nodule grows while on suppression therapy, this finding would be consistent with malignancy and would appear to be an indication for surgical excision7. Recent studies reveal otherwise. TSH appears to be one of many potential growth factors for thyroid tissue. The literature further documents that many benign nodules do not respond to suppressive therapy, while 13-15% of thyroid cancers may become smaller with suppressive doses of levothyroxine. The use of thyroid hormone suppression to separate benign from malignant disease should, therefore, be discouraged.4,7

ULTRASONOGRAPHY Ultrasonography is able to help categorize nodules as solid, cystic, or mixed solid and cystic with 90+ % accuracy.13,20 Ultrasound provides greater anatomic detail than thyroid scintigraphy or computerized tomography and is able to resolve cystic lesions as small as 2 mm and solid lesions measuring 3 to 4 mm in size. ULTRASONOGRAPHY ADVANTAGES

DISADVANTAGES

•

Noninvasive

•

May miss very small nodules

•

No radiation involved

•

•

Differentiates cystic from solid lesions

Unable to identify substernal goiter

•

Unable to differentiate malignant

•

Determines exact size

•

Can be used to guide needle aspiration

from benign tissue

Figure 4. Adapted from Dwarakanathan, in Comprehensive Therapy, 19935.

It has been thought that solid lesions carry a higher risk of malignancy than do cystic lesions. Recent data suggests that pure cystic lesions are rare, comprising less than one percent of all lesions and cysts frequently represent degenerative changes of solid lesions resulting in a cystic component. The frequency of malignancy is higher in solid lesions, but demonstration that a lesion has a cystic component does not eliminate its potential for malignancy (13-20% vs. 0.510% respectively, depending upon the series).19 Ultrasonography is relatively insensitive with regards to differentiating benign from malignant nodules/lesions.13 In a recent review of published series, 69% of all nodules were solid, 19% were cystic, and 12% were mixed. About 20% of solid nodules were found to be malignant while some 10% of mixed nodules were also noted to be malignant at histological examination.7,20 Most investigators would agree that the routine use of ultrasound in the evaluation of thyroid nodules is not cost effective. Ultrasonography does have a potential role in a number of areas. As stated previously, ultrasonography is useful in classifying lesions as solid, cystic or mixed. It also may be used to evaluate the size and volume of a thyroid nodule,13 as well as multiplicity of nodules in a gland clinically suspected to have solitary nodule, as 20-40% of patients referred for ultrasound of a solitary nodule will demonstrate one or more additional nodules.19 Ultrasonography is also useful for follow up evaluation of cystic nodules after aspiration of cyst contents and FNAB. If a residual palpable lesion is present following aspiration, ultrasonography can identify it as persistent cyst or solid nodule. A solid nodule should be sampled by FNAB; a cystic lesion that reaccumulates following aspiration may have an increased risk of malignancy (10%) compared to that of completely aspirated cyst (1%), and should be reaspirated or surgically excised.19 Ultrasonography can also be used: (1) to accurately assess a lesion over time for evidence of change in nodule size; (2) to preoperatively identify lymph nodes following FNAB diagnosis of malignancy; (3) to monitor the thyroid bed for recurrence following thyroidectomy for malignancy; and (4) to guide repeat FNAB when results initially are nondiagnostic, with small lesions, or those situated in difficult positions to blindly access.9,19

COMPUTERIZED TOMOGRAPHY AND MAGNETIC RESONANCE IMAGING Computed tomography and magnetic resonance imaging may occasionally be useful in identifying the extent and location of thyroid masses. The imaging studies are generally reserved for substernal or retrosternal thyroid masses, detection and delineation of thyroid nodules, and identification of regional or distant metastasis.9 These studies are not cost effective with regard to the initial management of thyroid nodules.

FINE NEEDLE ASPIRATION BIOPSY It is relatively difficult to overstate the influence of FNAB on the management of nodular thyroid disease. The procedure was first utilized in the 1950’s, but did not gain general acceptance until the 1980’s.22 During the last decade, FNAB has become the most direct, and accurate diagnostic procedure in the management of nodular thyroid disease.6,9 Most centers which have effectively utilized FNAB have demonstrated a 35-75% reduction in the number of patients requiring surgery, a doubling of the malignancy yield at thyroidectomy, and up to a 25% reduction in the costs associated with thyroid nodule management.4 FINE NEEDLE ASPIRATION BIOPSY ADVANTAGES

DISADVANTAGES

•

Safe, simple, quick office procedure

•

•

Yields direct cytologic information

Unable to distinguish benign from malignant follicular and Hurthle cell neoplasms

•

Will diagnose certain cancers (papillary, medullary, and anaplastic), as well as Hashimoto’s thyroiditis, and subacute thyroiditis with reasonable certainty

•

Complications (although infrequent) of invasive procedure

•

Requires experienced pathologist

•

Can be repeated many times

•

False negative diagnoses

•

Cysts can be diagnosed and cured

•

Avoids unnecessary surgery

Figure 5.. Adapted from Dwarakanathan, in Comprehensive Therapy, 19935.

THE PROCEDURE FNAB is generally performed as an outpatient procedure with the patient in the supine position with his or her neck flexed backward slightly. The skin is cleansed with alcohol. Generally local anesthesia is not required. A 23-25 gauge needle attached to a 10 ml disposable syringe is then place into the nodule after localization, and suction is applied while the needle is moved back and forth within the nodule. Suction is then released and the needle withdrawn from the nodule. The aspirated material is then expelled onto glass slides and either air dried and stained using the May-Grü nwald-Giemsa technique or immediately fixed with 95% alcohol and stained using a modified Papanicolaou method.6 Generally two to four aspirations are performed, depending upon the quality of the smears and the aspirationist.

CYTODIAGNOSIS

In order to allow for satisfactory categorization of FNAB material, the smear must contain an adequate number of cells. The real question here is, what is adequate? Various authors have published recommendations for determining the adequacy of FNAB smears. However, no standardized criteria exist for judging aspirate sufficiency. Gharib and Goellner have suggested that an adequate smear is one containing 5 or 6 groups of well preserved cells, each group containing 10-15 cells, minimum6. Hamburger has defined a satisfactory benign aspirate as having at least six clusters of benign appearing follicular epithelial cells on at least two slides prepared from separate aspirates. Any malignant appearing cell automatically dictates a cytodiagnosis of malignant. Specimens which do not meet these criteria are classified as nondiagnostic (rather than benign) 11. Standardization of cytodiagnostic criteria for thyroid FNAB can only serve to improve the accuracy of the technique. Cytodiagnosis of aspirates is divided into one of four categories on the basis of cellular appearance: benign, suspicious or indeterminate, malignant, and nondiagnostic.6 Accuracy of cytodiagnosis is influenced by the experience of the aspirator and the cytopathologist interpreting the smear.19

BENIGN LESIONS Patients with a benign cytodiagnosis do not have malignancy, but rather may have a normal thyroid gland, a colloid nodule, lymphocytic thyroiditis, subacute thyroiditis or another benign condition.6 Most benign smears demonstrate abundant colloid material and typical follicular cells. Approximately 70% (range 53-90%) of all nodules sampled will demonstrate benign cytodiagnostic results.

MALIGNANT LESIONS Patients with malignant cytodiagnostic results have findings indicating malignant cells consistent with primary or metastatic thyroid carcinoma including papillary carcinoma, medullary carcinoma, anaplastic carcinoma, lymphoma, or metastatic tumor.22 Rarely, a diagnosis of follicular or Hü rthle cell carcinoma is made on the basis of thyroid FNAB. Approximately 3.5% (range 1-10%) of all nodules will yield malignant results.6

SUSPICIOUS/INDETERMINATE LESIONS Patients with a suspicious or indeterminate cytodiagnosis on FNAB have specimens which demonstrate hypercellularity and a pattern suggestive of, but not diagnostic, for malignancy.6 Frequently, these lesions are follicular or Hü rthle cell neoplasms, although papillary lesions are sometimes noted; these lesions frequently contain little colloid material.19 Follicular and Hü rthle cell neoplasms may be benign adenomas or malignant carcinomas. The distinguishing characteristics include evidence for capsular or vascular invasion. FNAB rarely provides enough information to classify these lesions correctly.

Approximately 10 - 15% of all cytologic specimens are labeled as suspicious for malignancy or indeterminate. These usually are the result of the presence of follicular or Hü rthle cell neoplasms, and about 20% of these lesions are actually malignant.9,19 This number approaches 60% when lesions suggestive of papillary carcinoma are considered.9

NONDIAGNOSTIC LESIONS Nondiagnostic results are obtained when relatively few follicular cells are obtained by FNAB. This occurs in approximately 15% of all aspiration attempts, but varies depending on the experience of the aspirationist, the size and nature of the lesion, vascularity, and when the cytopathologist has set stringent diagnostic criteria for a satisfactory smear.9 Most commonly, nondiagnostic results are due to aspiration of hemorrhagic or cystic lesions (which constitute approximately 15-20% of all nodules) with resultant dilution of the specimen. The finding of a cystic lesion, however, should not dissuade the physician from aggressively pursuing the diagnosis as up to 14% of cystic lesions may be malignant.9 In the event of nondiagnostic results, repeat aspiration will usually yield a satisfactory specimen in about 50% of cases. Even with repetitive aspirations, a residual 10% of specimens will remain unsatisfactory for interpretation and are labeled as persistently nondiagnostic.6 It is important to understand that a nondiagnostic result does not imply a benign specimen.19

RESULTS OF FINE NEEDLE ASPIRATION BIOPSY Data from seven, large series of patients undergoing FNAB revealed that 18,183 specimens were obtained and examined. The rate of benign cytodiagnosis was 53-90% (mean 69%), malignant cytodiagnostic findings occurred in 1-10% (mean 3.5%). Suspicious or indeterminate findings were present in 5-23% (mean 10%) and nondiagnostic specimens comprised 2-21% (mean 17%) of all specimens. Caruso and Mazzaferri recently reported similar findings in an analysis of 10 series containing 9119 patients -- 74% benign, 4% malignant, and 22% suspicious; 10% were nondiagnostic.6 Comparison of Diagnostic Cytologic Categories in Seven Series SERIES

COUNTR Y

YEAR

TOTAL CASES

BENIGN

MALIGNA NT

SUSPICIO US

NONDIAGNO STIC

n (%)

n (%)

n (%)

n (%)

n (%)

Gardiner

Canada

1986

1 465

1085 (74)

16 (1)

146 (10)

218 (15)

Hawkins

Spain

1987

1 399

1253 (90)

56(4)

68 (5)

22(2)

Khafagi

Australia

1988

618

404 (65)

29 (5)

68 (11)

117 (19)

Hall

US

1989

795

509 (64)

81 (10)

75 (9)

130 (16)

Altavilla

Italy

1990

2 433

1890 (78)

32 (1)

119 (5)

392 (16)

Caplan

US

1991

502

268 (53)

24 (5)

116 (23)

94 (19)

Gharib/ Goellner

US

1991

10 971

7071 (64)

416 (4)

1192 (11)

2292 (21)

18 183

12 480 (69)

654 (3.5)

1784 (10)

3265 (17)

TOTAL

Figure 6. Adapted from Gharib and Goellner, in Annals of Internal Medicine, 19936.

FALSE DIAGNOSES FALSE NEGATIVE DIAGNOSES False negative results are worrisome because they imply missed malignant lesions. They occur secondary to sampling error or interpretive mistakes. The false negative rate is defined as the percentage of patients with benign cytologic findings, who are subsequently found to have malignant lesions of the thyroid. Therefore, the true false negative rate could be found only in a series in which all patients with benign results screened by FNAB, had subsequent surgery with histologic review. No such study has been performed. The reported false negative rates in the seven series summarized ranged from 1.3%-11.5% with an average of 5.2%. When data from the series are combined, only 10% of patients with benign cytologic findings underwent thyroid surgery. Caruso and Mazzaferri found a false negative rate of 5% (range 1-6%), based on their pooled data from 10 series in which only 14% of nodules were excised. Campbell and Pillsbury, analyzing data from 912 patients with benign cytologic results and subsequent histologic examination found a pooled false negative rate of 2.4%, with a range of 0.5-11.5%. Ashcraft and Van Herle in their analysis of 1330 patients found a false negative rate of 1.7%.6 Boey and colleagues took a different approach in 365 patients with negative/benign cytologic results and followed them for a period of 30 months, during which time they found only 2 cancers - a false negative rate of 2.1%. Grant and colleagues similarly followed 439 patients who had benign results on FNAB but did not undergo immediate surgery -- 3 patients subsequently demonstrated malignancy in a 6.1 year follow up period, for a false negative rate of 0.7%.6 Most authorities agree that the true false negative rate probably lies below 5% but an accurate value may never be realized.6

FALSE POSITIVE DIAGNOSIS The false positive rate is the percentage of patients with malignant results on FNAB who are found to have benign lesions at surgery. In Gharib and Goellner’s review, the false positive rate varied from 0-7% (mean 2.9%). These results were similar to the rates of 0.5% reported by Ashcraft and Van Herle, 1.2% reported by Campbell and Pillsbury and 6% reported by Caruso and Mazzaferri.6

SPECIFICITY AND SENSITIVITY Two values used to estimate the accuracy of FNAB in the evaluation of thyroid nodules for malignancy are sensitivity and specificity. The estimates of sensitivity and specificity depend on how the suspicious/indeterminate category is handled. If the suspicious results are considered positive, the sensitivity increases and specificity decreases (98% sensitive, 72% specific). If the suspicious results are considered negative, then specificity increases and sensitivity decreases (65% sensitive, 100% specific). An overall sensitivity of 83% and specificity of 92%, as quoted by Gharib, translates into a diagnostic accuracy greater than 95%, thus confirming FNAB as a reliable test.6

POSITIVE AND NEGATIVE PREDICTIVE VALUES According to the data presented by Gharib and Goellner, and a review by La Rosa, a malignant result on FNAB has a positive predictive value of about 96%, while a benign result suggests a negative predictive value of 99%6,12. PRECISION OF FINE NEEDLE ASPIRATION BIOPSY series

nodules operated

malignant

false neg

fals e

sensitivity

specificit y

accuracy

pos

gharib and

1750

682 (39%)

2.0%

0 .7%

98%

99%

98.5%

La ROSA

827

250 (30%)

1.7%

2 .2%

98%

98%

97.7%

TOTAL

2577

932 (36%)

goellner

Figure 7. Adapted from Giuffrida and Gharib, in American Journal of Medicine, 19956.

LIMITATIONS FNAB has two major limitations : nondiagnostic results, and suspicious or indeterminate results. Nondiagnostic results often occur in the setting of cystic or vascular lesions. Repeat biopsy may produce satisfactory results in about half of the patients. The success of FNAB depends upon careful aspiration, smear preparation, and cytodiagnostic interpretation. Realize that approximately 10% of all nodules will yield nondiagnostic results regardless of the number of aspirations performed. The second limitation is the indeterminate result. Here, 20-30% of aspirations show suspicious

cytodiagnosis because of the difficulty differentiating benign cellular follicular and Hurthle cell neoplasms from their malignant counterparts. About 25% of suspicious lesions are found to be malignant at surgery. Methods of more accurately diagnosing these malignant lesions preoperatively have been suggested, but have not significantly improved cytodiagnostic yield.6 Thyroid scintigraphy may be of benefit in identifying functioning nodules which have a lower, but not negligible risk for malignancy, and thus avoiding surgical intervention.6

EFFECT ON THE MANAGEMENT OF THYROID NODULES Many studies have shown that FNAB has had a significant effect on the number of patients proceeding to surgery, and upon the incidence of carcinoma found at operation. Miller and colleagues reported that the availability of biopsy decreased the percentage of patients undergoing thyroid surgery from 48% to 24% while increasing the yield of malignancy from 12% to 14%. Hamburger reported that following the introduction of FNAB, the percentage of patients undergoing surgery decreased from 67% to 43%, with the percentage of carcinomas found in operated patients doubling from 14% to 29%. Caplan and colleagues documented similar results, with the percentage of patients proceeding to thyroidectomy decreasing from 61% to 33%, while the yield of malignancy doubled from 18% to 39%. Alternatively, Gharib found that FNAB resulted in a 12% increase in the frequency of surgery for benign follicular and Hurthle cell neoplasms/adenomas, but 25% fewer thyroid operations were performed for benign disease.6 The economic effect that FNAB has produced is considerable and has resulted in a significant reduction in the overall cost of management by as much as 25% in some series. The main effect in cost reduction has been the elimination of unnecessary thyroid surgery in patients with benign lesions.6

APPROACH TO THE PATIENT WITH A THYROID NODULE Most patients are asymptomatic when thyroid nodules are first identified, as most are discovered incidentally. Patients will, however, present with complaints of "lumps in their neck". It is of utmost importance to perform a careful history and physical exam, with particular attention to risk factors for thyroid malignancy.21 Specifically, the patient should be questioned about previous history of thyroid disease, head, neck, or chest irradiation, the pattern of progression of the nodule, as well as family history of thyroid carcinoma. The patient should be thoroughly questioned regarding symptoms of compression, obstruction, or invasion including dyspnea, dysphagia, hoarseness, and pain. Physical exam should focus on identifying nodular disease in the thyroid, characterizing the thyroid with regards to the number, quality, and distribution of nodules, gland size, texture, and presence of lymphadenopathy.4 Effective management of the patient with a thyroid nodule revolves around an assessment of their risk for malignancy. Although the aforementioned risk factors, as well as age and sex of the patient, lack specificity and sensitivity, these risk factors may influence the management decisions involving a highly suspicious nodule. Patients with

multiple risk factors (high risk) for malignancy may most safely be managed with thyroidectomy in spite of FNAB cytodiagnosis. There is considerable diversity and controversy in the approach to the patient with a thyroid nodule. Some authors favor early surgical approach to patients with thyroid nodularity as they argue that in as many as 10% of patients with thyroid nodules, histology may reveal malignancy which is a potentially lethal condition dictating aggressive treatment. Those favoring a more conservative approach argue that thyroid nodularity is an extremely common condition, of which a relative few number of patients will have thyroid carcinoma. An aggressive approach therefore may not be warranted or justified. Additionally, certain of the thyroid carcinomas are so indolent and slow growing that most patients will die of causes unrelated to the thyroid cancers.10 Alternative approaches to immediate surgery for all nodules include use of one or more of the following methods: radionuclide imaging, ultrasound, FNAB and trial of thyroid hormone suppression.4 As preliminary studies, thyroid function testing is generally normal. Minor abnormalities of thyroglobulin may be present, but this is a nonspecific finding. A suppressed TSH indicates likely hyperthyroidism, and if a palpable nodule corresponds to a hot/hyperfunctioning nodule on scintigraphy, the likelihood of malignancy is reduced (but not eliminated) and the patient can be treated with either radioablative therapy or surgical excision (subtotal thyroidectomy). Many centers still obtain both thyroid scintigraphy and ultrasound as the initial diagnostic evaluations in a patient with a thyroid nodule. Other centers, based upon its excellent accuracy and reliability, utilize FNAB to obtain tissue cytology as the primary step in evaluation of thyroid nodules.19 Either approach is reasonable. The deciding factors should include the availability of resources, patient and physician preference, and the overall clinical assessment. The availability of a physician with adequate expertise in obtaining FNAB specimens, as well as an experienced cytopathologist familiar with the interpretation of thyroid aspirates cannot be overemphasized when making a management decision.19 Thyroid ultrasound and scintigraphy, with their known inability to accurately and reliably differentiate benign from malignant disease, may be appropriate studies when the patient is extremely anxious regarding FNAB, when no clinical risk factors for malignancy are present, when clinical signs of toxicity are suspected or present, when the patient is anticoagulated (although this poses little risk when FNAB is performed by an experienced operator), or when an experienced aspirationist or cytopathologist is not available.19,20 FNAB, with its high degree of accuracy and reliability, ease of performing, and general acceptance, is likely the most reasonable initial diagnostic procedure in the evaluation of nodular thyroid disease. As this approach is readily available at this institution, further discussion will be primarily based upon the approach utilizing FNAB as the major diagnostic tool. If FNAB yields a malignant cytodiagnosis, near total thyroidectomy (with modified neck dissection if lymph node metastases are suspected) is the procedure of choice.4,10,19,22 Nodules which yield suspicious or indeterminate cytodiagnosis on FNAB present a special problem. No clinical or laboratory finding will accurately predict which patients with suspicious cytologic results, will in actuality have a malignant process.6 As stated previously, 10-30% of nodules with suspicious cytodiagnoses will yield malignancy on histologic diagnosis.9 Typically these lesions

represent follicular or Hü rthle cell neoplasms, which cannot be adequately characterized by FNAB alone, because capsular or vascular invasion cannot be seen. Nodules which are labeled as suspicious for papillary thyroid carcinoma were found in one study to be malignant in about 60% of cases at thyroidectomy.4 It is for these reasons that most authors recommend surgical excision with lobectomy and frozen section analysis at operation to determine whether further surgical excision/dissection need to be attempted. This is particularly emphasized if the patient demonstrates risk factors making the likelihood of finding malignancy higher -- a male patient, older or younger age group, suspicious characteristics, or a lesion larger than 3 cm.11 Some authors have recommended that following a diagnosis of suspicious or indeterminate, radionuclide imaging be performed. If the nodule is cold/hypofunctional, surgery is subsequently recommended. If the patient is euthyroid and the nodule functional, observation can be safely undertaken.6 A hot nodule in a patient with clinical or biochemical evidence of thyrotoxicosis could be treated with intranodular ethanol injection, radioiodine, or surgery.10 Still, even with this approach, many patients will proceed to surgery, and potential malignancies may be missed with conservative observation of hot nodules. Thyroid suppression therapy may also be employed. In a lesion noted to be benign by cytodiagnosis, most authors recommend careful follow up with periodic exams and repeat aspiration in one year if necessary.10,19 There is currently much debate regarding whether patients with a benign thyroid nodule should be followed with simple observation or whether levothyroxine suppression therapy should be administered.6 This issue will be discussed further in a later section of this paper. Nondiagnostic cytodiagnostic results are encountered in about 15% of all FNAB attempts. This is usually due to the presence of cystic or hemorrhagic nodular material and the acquisition of an insufficient number of thyroid follicular cells on which to make a cytodiagnosis.6 Reaspiration of a solid nodule will yield a diagnostic specimen in about 50% of these cases, from which point appropriate management decisions can be made. In the event that a cytodiagnosis is not obtained, many authors recommend surgical excision with lobectomy or nodulectomy and frozen section analysis to determine the need for further surgical intervention.4 In the case of cystic lesions, often the initial aspiration will cure the nodule, even though it does not provide cytodiagnosis. If after initial aspiration of a cyst, a palpable nodule remains, FNAB should again be performed on the residual tissue. Decompressed nodules that recur can be reaspirated, but many authors recommend surgical excision, particularly if the lesion is larger than 3 cm, as these lesions carry an increased risk of malignancy.4,6

SPECIAL CONSIDERATIONS AUTONOMOUSLY FUNCTIONING THYROID NODULES Autonomously functioning thyroid nodules (as identified on radionuclide scanning) often grow insidiously and may be present for many years before overt thyrotoxicosis occurs. These nodules function independent of TSH stimulation, and therefore should not be treated with thyroid hormone in an attempt to suppress the tissue. Autonomously functioning tissue may be treated

with surgery, radioablative therapy, or intranodular ethanol injection. Surgical resection is used frequently at some centers due to concerns about possible side effects of radioactive iodine therapy, availability of tissue to rule out malignancy, and the little time required to achieve a euthyroid state. Intranodular ethanol injection is a relatively new technique that has been used effectively in several European centers with significant shrinkage of the nodule achieved, and with moderate control of hyperthyroidism. The technique has been associated with transient vocal cord paralysis, exacerbation of hyperthyroidism, and persistent subclinical hyperthyroidism in a significant number of cases.4 No long term follow up relating to the efficacy of this technique is currently available. The management of the clinically euthyroid patient with a hot nodule is also controversial. In a clinically euthyroid patient with a hot nodule less than 2.5 cm in diameter, observation is reasonable. In patients with hot nodules greater than 2.5 cm in diameter with or without toxicity, the presence of increased risks for osteoporosis or cardiovascular disease, and in an elderly patient, definitive therapy with surgery or 131I radioablation is appropriate.18 Patients with nontoxic hot nodules can be followed clinically. Serial exams and thyroid function testing should be performed at regular intervals to ensure that the patient, particularly the elderly who often have less symptoms than younger patients, are not thyrotoxic.18

OCCULT NODULAR THYROID DISEASE As implied previously, clinical examination of the thyroid by palpation is relatively insensitive in terms of identifying nodular thyroid disease. Mortensen, in an autopsy series in the 1950’s, reported that one or more thyroid nodules were detected in 50% of patients whose glands appeared clinically normal. In as recent publication by Ezzat et al, palpable nodules were identified in 21% of his "normal" patient population (9% single/solitary, and 12% multiple nodules). In contrast, 67% of these patients had one or more clinically unappreciated nodules by ultrasonography (22% solitary, 45% multiple).9 Occult lesions of the thyroid, defined as impalpable nodules usually smaller than 2.0 cm, are a management dilemma for physicians. Occult papillary thyroid cancers less than 1cm are frequently found at autopsy and in surgical specimens. These tiny papillary cancers apparently have, and are suspected of having, little clinical relevance as they are never manifest during life.14 The incidental ultrasonographic finding of a nodule should not automatically require further investigation. This is impractical, and there is no data on the malignant potential of such nodules. For lesions less than 2.0 cm, a negative family history of thyroid carcinoma, and no history of head or neck irradiation, observation of the nodule is reasonable. For nodules greater than 2.0 cm, a positive family history, presence of other risk factors suggestive of malignancy, or ultrasongraphic characteristics suggestive of malignancy, further diagnostic evaluation with biopsy (ultrasound guided if necessary) is warranted. Malignant or suspicious cytodiagnostic findings should be managed surgically as noted previously.9

MULTINODULAR GOITERS

Multinodular goiters have different nodule sizes and variable radionuclide uptake, producing irregular contour and multiple nodules that appear hotter or colder than the remainder of the gland. Multinodular goiters are common in iodine deficient areas, but may also occur in iodine rich areas. The rate of nodular growth is usually slow, but varies from person to person and nodule to nodule, as some nodules are more sensitive to TSH and other growth factors.18 With regards to the risk for malignancy, Belfiore and associates studied a group of 4,485 patients with solitary thyroid nodules along with a group of 1,152 patients with multiple nodules and demonstrated that the cancer rate was similar (approximately 4%) between the two groups.1 The general, and currently accepted theory, is that multinodular goiters have a lower malignancy potential. This new information may contradict that long-standing belief, and all patients with thyroid masses must be considered for possible malignancy, especially when a dominant nodule is noted. Further study in this area is warranted. Again, patients with risk factors for cancer must be considered for possible malignancy, regardless of the type of nodule or goiter with which they present.

GRAVES’ DISEASE Patients with Graves’ disease who are shown by thyroid scintiscanning to harbor a cold or hypofunctional nodule in a diffusely enlarged thyroid gland carry an increased likelihood of malignancy, which has previously been demonstrated to be more aggressive than that found in euthyroid patients. Surgical referral for near total thyroidectomy is recommended.22

Thyroid Hormone Suppressive Therapy Long term levothyroxine suppression therapy to arrest growth or shrink thyroid nodules is a practice that remains controversial. The use of thyroid hormone suppression is based upon the belief that nodule growth is TSH dependent.4,9 There is no well documented correlation between response to therapy and variables such as age, duration of nodule presence, size, or nodule type.9 In the past 30 years, there have been more than 20 studies examining the efficacy of thyroid hormone suppressive therapy on thyroid nodules. Early studies were generally uncontrolled analyses flawed by an inability to document adequate TSH suppression or document nodule change accurately.4 In the past 10 years, nine studies have been performed to evaluate this issue further including five randomized trials and four uncontrolled analyses. These more recent studies benefit from the use of improved TSH assays to verify adequate suppression, FNAB to exclude malignant or suspicious lesions, and high resolution ultrasound for accurate nodule sizing.4 The four nonrandomized trials of suppressive therapy for thyroid nodules included a total of 637 patients; 431 of these patients had previously received radiation to the head and neck. Three studies of nonirradiated patients, utilizing second or third generation TSH assays and high resolution ultrasound to assess nodule volume change, demonstrated response rates of 34-56%. The basic fault of these studies is a failure to distinguish spontaneous regression from therapeutic response. In randomized, placebo controlled trials, rates of spontaneous regression have ranged

from 1-35%.4 The five randomized trails performed include a total of 159 patients receiving thyroid hormone and 157 untreated patients. Utilizing sensitive TSH assays and high resolution ultrasound, only a single study found mean nodule volume to decrease in treated patients compared to untreated controls. However, two of the four studies, with available individual patient data, demonstrated a subgroup of patients deemed thyroxine responsive --26% of treated patients experienced significant nodule regression compared to 15% of untreated control patients.4 In a recent randomized controlled study by La Rosa, thyroid hormone suppression therapy was demonstrated to be effective in reducing nodular volume in 39% of treated patients compared to 0% of untreated patients. Moreover, following discontinuation of therapy, 25% of patients previously treated with thyroid hormone demonstrated clinically significant increases in nodule volume. In patients previously untreated who then received thyroxine, nodule volume was subsequently shown to decrease.12

Side Effects of Thyroid Suppression Therapy EFFECTS ON CARDIAC FUNCTION The cardiac effects of overt hyperthyroidism have been described in detail and may be grouped into three basic categories: (1) direct effects on cardiac myocyte protein synthesis, size and intracellular calcium handling; (2) effects on peripheral hemodynamics and oxygen utilization; and (3) sympathoadrenal effects. The clinical correlates of these effects include an acceleration of the pulse rate, an increase in systolic blood pressure, an increase in the stroke volume and cardiac output, a decrease in the systemic vascular resistance, an increase in cardiac muscle mass, and a tendency toward atrial dysrhythmia. Correction of overt thyrotoxicosis results in a normalization of these effects4. Several recent studies have extended these observations to patients with subclinical hyperthyroidism, defined as a suppression of TSH due to an increase in serum thyroid hormone levels within the confines of the normal range4. The most detailed work to date in this area has evaluated the effects of subclinical hyperthyroidism on cardiac mass and chamber size, systolic time intervals, resting pulse rate, arrhythmia, and symptom rating scale, as well as an assessment of the effects of beta adrenergic blockade on each of these parameters. Patients undergoing thyroid hormone suppressive therapy at doses sufficient to suppress serum TSH levels had a significant elevation in resting pulse, an enhanced rate of atrial dysrhythmia, an increased interventricular septal and posterior wall thickness, increased left ventricular mass, and an enhancement in indices of systolic function when compared with age and sex matched controls. A similar group of patients undergoing thyroid suppression therapy experienced improvements in a hyperthyroid symptom rating scale, decreases in heart rate and incidence of atrial dysrhythmia, and a return of cardiac muscle mass to normal values when given a 6 month course of beta blockers in addition to thyroid hormone. These studies suggest that a wide spectrum of cardiac abnormalities may be present in patients undergoing thyroid hormone suppressive therapy, and suggest a means of alleviating these effects (i.e. beta blockers) in patients in whom suppressive therapy is mandated by clinical

circumstances4. Atrial fibrillation is a well recognized complication of overt hyperthyroidism. The clinical significance of atrial fibrillation is evidenced by the dire consequences of this arrhythmia, such as congestive heart failure, systemic embolism, and stroke. Occult thyrotoxicosis was found in 13 of 75 consecutive patients presenting with atrial fibrillation and no identifiable cardiac cause4.

EFFECT ON BONE MINERAL DENSITY Thyroid hormone is known to enhance osteoclastic bone reabsorption, an effect that appears to be mediated indirectly through effects of T3 upon osteoblasts, which induce osteoclastic activity. Studies involving skeletal effects of exogenous thyroid hormone and subclinical hyperthyroidism in premenopausal women have yielded varying results, with either no difference from control subjects, or diminished bone density in the hip, forearm, and lumbar spine. A recent meta analysis involving premenopausal women, including 441 measurements of bone density in 239 patients from nine separate studies, led to a composite premenopausal woman, 39.6 years of age, taking 164 mcg/d of levothyroxine for 8.5 years and with a theoretical bone structure consisting of 30% distal forearm, 29% femoral neck, and 41% lumbar spine. This composite woman would be expected to experience an excess bone mass loss of 2.7% over 8.5 years, not significantly different from that in control subjects4. The data are somewhat more incriminating in studies involving postmenopausal women, although no well designed, long term, prospective studies have been performed to date. Several studies have found significantly lower bone density in the hip, spine, and forearm in postmenopausal women with subclinical hyperthyroidism or taking thyroid hormone when compared with control subjects. A meta analysis involving postmenopausal women including 317 bone density measurements in 149 patients comprising eight different studies gave a composite patient 61.2 years of age, treated with 171 mcg/d of levothyroxine for 9.9 years and with a bone structure consisting of 42% femoral neck, 47% lumbar spine, and 11% distal forearm. This aggregate patient would be expected to experience a significant excess bone loss totaling 9% over 9.9 years of thyroid hormone therapy. Another study noted that women taking estrogen, in addition to thyroid hormone therapy, experience less bone loss than those taking thyroid hormone alone. It is likely that a spectrum of bone pathology occurs depending on the degree and duration of subclinical thyrotoxicosis as well as the presence of other known risk factors for osteoporosis4.

Recommendations Regarding Thyroid Hormone Suppressive Therapy Given that the majority of patients with thyroid nodules do not experience nodule shrinkage with suppressive therapy and in light of rising concern over the potential adverse effects associated with subclinical hyperthyroidism, it is difficult to recommend the indiscriminate utilization of thyroid hormone suppressive therapy in all patients with benign thyroid nodules.4 For those patients in whom thyroid hormone suppressive therapy is deemed appropriate and attempted, a limited trial of suppressive therapy with careful follow up evaluation of nodule size

by high resolution ultrasonography and assessment of TSH is reasonable. A TSH level of 0.1mIU/L - 0.3 mIU/L should be the goal of therapy, as there has been little data demonstrating increased suppression of thyroid nodules at lower levels.4 In postmenopausal women and in patients with a history of cardiac disease, many authors would argue instituting therapy only if the nodule enlarges during a 6-12 month period of observation. In postmenopausal women and premenopausal women at significant risk of osteoporosis, suppressive therapy should be accompanied by bisphosphonate or estrogen therapy, or both.9 Apparent responders to thyroid hormone suppressive therapy should be followed closely and therapy discontinued after the trial of suppression therapy to distinguish spontaneous regression from therapeutic effect. Therapy is reinstituted in those patients whose nodules regrow following discontinuation of therapy.9 If the nodule fails to shrink, the medication should be discontinued. Conversely, if the nodule enlarges during therapy, repeat FNAB or surgery is advised.17

CONCLUSION A tremendous effort by a number of authors has been put forth to elucidate the most efficient, cost effective, and accurate method of evaluating and managing nodular thyroid disease. The overwhelming volume of literature concerning this topic can easily confuse the clinician who is faced with a patient demonstrating a thyroid nodule. One of the most important points to remember when choosing the best method of evaluating a patient with a thyroid nodule is that the clinical situation, assessment of risk for malignancy, and available resources must be taken into account. Thyroid nodules are extremely common and frequently benign. For the most accurate diagnostic approach, fine needle aspiration biopsy should be employed. Alternative imaging techniques, such as nuclear thyroid scanning and ultrasonography have limited clinical utility and are highly unreliable for differentiating benign from malignant disease. Nodules with a malignant and suspicious cytodiagnosis of FNAB should be managed with referral to an experienced endocrine surgeon for thyroidectomy. Malignant nodules should be managed with near total thyroidectomy, and suspicious nodules should be treated with lobectomy with frozen section analysis to determine if further excision is required. Benign nodules on cytology can be safely managed with observation. Nondiagnostic cytology should result in repeat aspiration and appropriate treatment when diagnostic cytology is obtained. The nodule which retains a nondiagnostic cytology should be excised. Thyroid hormone suppression therapy has become a highly controversial issue in recent years, and has been associated with adverse effects on both the cardiac and skeletal systems. The use of levothyroxine suppression therapy should be undertaken with caution and must be closely monitored.

ACKNOWLEDGEMENTS I am deeply grateful to Drs. K. Patrick Ober, and William T. Cefalu for their review and criticisms of this manuscript, and my wife, Alicia, who reviewed the manuscript on numerous occasions, as well as tolerated and supported me during the stressful preparation of the manuscript and presentation.

BIBLIOGRAPHY 1. Belfiore, A., et al. The Management oof Thyroid Nodules. Journal of Endocrinologic Investigation. 1995; 18: 155-58. 2. Belfiore, Antonio, et al. Cancer Risk Prevalence in Patients with Cold Thyroid Nodules : Relevance of Iodine Intake, Sex, Age, and Multinodularity. American Journal of Medicine. Oct 1992; 93: 363-69. 3. Blum, Manfred. Why Do Clinicians Continue to Debate the Use of Levothyroxine in the Diagnosis and Management of Thyroid Nodules. Annals of Internal Medicine. Jan 1, 1995; 122(1): 63-4. 4. Burch, Henry B. Evaluation and Management of the Solid Thyroid Nodule. Endocrinology and Metabolism Clinics of North America. Dec 1995; 24(4): 663-710. 5. Dwarakanthan, Arcot A. Evaluation of Thyroid Nodules. Comprehensive Therapy. 1993; 19(5): 220-24. 6. Gharib, Hossein, and Goellner, John R. Fine Needle Aspiration Biopsy of the Thyroid: An Appraisal. Annals of Internal Medicine. 1993; 118: 282-89. 7. Gharib, Hossein. A Strategy for the Solitary Thyroid Nodule. Hospital Practice. Sept 30,1992; : 53-60. 8. Gharib, Hossein. Subspecialty Clinics : Endocrinology/Metabolism. Fine Needle Aspiration Biopsy of Thyroid Nodules : Advantages Limitations, and Effect. Mayo Clinic Proceedings. Jan 1994; 69: 44-49. 9. Giuffrida, Dario, and Gharib, Hossein. Controversies in the Management of Cold, Hot, and Occult Thyroid Nodules. American Journal of Medicine. Dec 1995; 99: 642-50. 10. Gupta, Krishan L. Neoplasm of the Thyroid Gland. Clinics in Geriatric Medicine. May 1995; 11(2): 271-290. 11. Hamburger, Joel I. Extensive Personal Experience. Diagnosis of Thyroid Nodules by Fine Needle Biopsy : Use and Abuse. Journal of Clinical Endocrinology / Metabolism. 1994; 79(2): 335-39. 12. La Rosa, Giacomo L., et al. Levothyroxine and Potassium Iodide Are Both Effective in Treating Benign Solitary Cold Nodules of the Thyroid. Annals of Internal Medicine. Jan 1, 1995; 122(1): 1-8.

13. Mazzaferri, Ernest L. Management of a Solitary Thyroid Nodule. Current Concepts. Feb 25,1993; 328(8): 553-59. 14. Mazzaferri, Ernest L. Thyroid Cancer in Thyroid Nodules : Finding a Needle in a Haystack. American Journal of Medicine. Oct 1992; 93: 359-62. 15. Molitch, Mark E. et al. The Cold Thyroid Nodule: An Analysis of Diagnostic and Therapeutic Options. Endocrine Reviews. 1984; 5(2): 185-199. 16. Reverter, J.L., et al. Suppressive Therapy with Levothyroxine for Solitary Thyroid Nodules. Clinical Endocrinology. 1992; 36: 25-28. 17. Rifat, Sami F., and Ruffin Mack T. Management of Thyroid Nodules. American Family Physician. Sept 15, 1994; 50(4): 785-90. 18. Rolla, Arturo R. Thyroid Nodules in the Elderly. Clinics in Geriatric Medicine. May 1995. 11(2): 259-69. 19. Ross, Douglas S. Evaluation of the Thyroid Nodule. Journal of Nuclear Medicine. 1991; 326: 2181-2192. 20. Sheppard, M.C., and Franklyn, J.A. Management of the Single Thyroid Nodule. Clincial Endocrinology. 1992; 37: 398-401. 21. Whitman, Eric D., and Norton Jeffrey A. Endocrine Surgical Diseases of Elderly Patients. Surgical Clinics of North America. Feb 1994; 74(1): 127-144. 22. Woeber, Kenneth. Cost Effective Evaluation of the Patient with a Thyroid Nodule. Endocrine Surgery. June 1995; 75(3): 357-363.