Thyroid carcinoma seems to have generated a considerable number of misconceptions. About 20% of these tumors are “pure” papillary, about 10% pure follicular, about 50% mixed papillary and follicular, and about 5% (range, 2%–10%) are called medullary. However, the pure papillary carcinoma usually has a few follicular elements if enough histologic sections are made, and the reverse is sometimes true in follicular tumors. In addition, some pathologists classify the tumors according to the predominant element unless the proportions of each element are very similar. If this were done, about 65% would be called papillary and about 20% follicular. There is enough diversity in classification methods to create difficulty in relating pathology reports to statistics in the literature. Papillary and most mixed papillary-follicular carcinomas metastasize primarily to regional lymph nodes. Prognosis is excellent in young adults but less so in older persons. Follicular carcinoma tends to produce hematogenous metastases, most often to lungs and bone. About 15% (range, 4%–30%) of single palpable nodules not selected by thyroid scan or fine-needle aspiration are malignant when excised.

Thyroid radionuclide scan. A major screening test is the thyroid scan. The characteristic appearance of thyroid carcinoma is a single nonfunctioning nodule. A gland that is multinodular on scan has less chance of containing carcinoma than one with a solitary nodule. On occasion, a palpable nodule may represent metastatic carcinoma from another primary site in a lymph node close to the thyroid.

Radionuclide scanning of thyroid nodules can be done with radioactive iodine (RAI) or technetium 99m pertechnetate. Results from comparison studies usually agree, but occasionally carcinomas that appear to have some function on technetium scan but not on iodine scan have been found. About 20% of single thyroid nodules without demonstrable function on scan are malignant (literature range, 3%–58%). About 6% of nodules with some function (reduced, but present) and about 6%–8% of nodules with apparent normal function (literature range, 0%–38%) are reported to be malignant. In some of these cases, normal thyroid tissue above or below the nodule creates a false impression of nodule function. Hyperactive nodules are very rarely malignant, although occasionally a malignancy is found unexpectedly in the same gland.

A minority of investigators believe that radioiodine or technetium scanning is not helpful in evaluation of thyroid nodules for possible malignancy. As noted previously, a single nodule without demonstrable function on scan has roughly a 20% chance of malignancy, which means that 80% of such nodules will be falsely positive for malignancy. On the other hand, some reports indicate that 6%–8% of nodules with apparently normal function may actually be malignant and thus represent false negative results. Therefore, some investigators rely on criteria other than thyroid scan to determine which patients with thyroid nodules should receive operative therapy. The criteria that have been used include patient history, characteristics of the nodule on physical examination, fine needle aspiration, or response of the nodule to thyroid hormone suppression. In the suppression test, failure of the nodule to diminish at least 50% in size during 3 months of suppression would increase the chance of malignancy.

A significant number of patients are referred for thyroid scan while thyroid uptake of radionuclide is being suppressed by administration of thyroid hormone or by x-ray contrast media. This frequently produces unsatisfactory or even misleading results.

Thyroid scan to detect thyroid carcinoma metastases. A different problem may arise when thyroid cancer is discovered and patients are referred for scanning to detect metastases, either before or after initial therapy. Unless all of the normal thyroid tissue is removed or is ablated by radioiodine therapy, enough of the scanning dose will be taken up by normal tissue to make such attempts useless in most cases. In addition, replacement thyroid hormone administration must cease for 2-4 weeks before scanning, so that the pituitary will once again produce thyroid-stimulating hormone (TSH), which, in turn, will help stimulate the tumor to take up the radioiodine. The dose of (RAI (1–5 mCi; SI, 0.037–0.185 MBq) for a metastatic tumor scan is more than 10 times the usual thyroid scan dose, and the optimal time to scan is 72 hours after administration of the dose. Some prefer a technetium phosphate bone scan to an iodine 131 (131 I) tumor scan. Most bone metastases detected by iodine are also detected by technetium phosphate, and the remaining thyroid tissue does not have to be ablated. However, a few bone metastases are detected by radioiodine and not by technetium. Lung metastases or recurrent neck tumor would be missed using technetium bone scan agents.

Serum thyroglobulin (TG) assay. Serum thyroglobulin (TG) assay has been advocated to follow patients after treatment of thyroid carcinoma. TG is synthesized by thyroid epithelial cells. It is present in measurable amounts in the serum of normal persons on immunoassay (using antibodies against TG) and is increased following TSH stimulation. Elevated values are found in active thyrotoxicosis (diffuse or nodular), thyroiditis, iodine deficiency, benign thyroid adenomas, and differentiated thyroid carcinomas. Therefore, TG elevation is too nonspecific to use for diagnosis of thyroid carcinoma. In thyroid carcinoma, the TG level is usually elevated in papillary, follicular, and mixed papillary-follicular neoplasms. Some anaplastic thyroid carcinomas produce elevated values and some do not. Medullary carcinomas do not produce measurable serum levels of TG. The TG assay can be used to monitor the progress of differentiated thyroid carcinomas after treatment. The half-life of circulating TG is said to be 8-22 hours, so circulating levels should be absent in 7-14 days after total destruction of all normal thyroid tissue and tumor tissue by surgery. Ablation by radioactive iodine is much more gradual and variable. TG values that are nondetectable or nearly so following therapy signify that no residual thyroid or tumor remains, and future elevations mean tumor recurrence or metastasis. Thyroglobulin values that are within the reference range following therapy could either be tumor or could be remnants of normal thyroid, and a thyroid tumor scan with 131 I is required to differentiate these possibilities.

One advantage of TG monitoring is less need for 131 I scanning. This avoids both additional radiation and the need to temporarily stop thyroxine replacement therapy to perform the scan. Also, occasionally patients with metastases associated with elevated TG levels but not detected on 131 I tumor scan have been reported. Disadvantages include a small number of patients with metastases detected on 131 I tumor scan but TG values within the reference range. This occurs in about 4% of cases (literature range, 0%–63%). TG levels are more likely to be normal with pure papillary tumors or those with only lung metastases. Also, the presence of patient anti-TG autoantibodies may interfere with the TG assay.

Fine-needle aspiration cytology. Fine-needle aspiration of thyroid nodules with cytologic smear examination of the aspirate has been advocated to aid diagnosis and, when possible, to replace surgical biopsy. Results in the literature vary rather widely, partially depending on experience with the technique, patient selection, and method of reporting positive results (for example, “definitely malignant” would detect fewer cases of carcinoma than the combination of “malignant” and “suspicious for malignancy”). Some centers report a false negative rate of less than 5% and a false positive rate of less than 2%. Most hospitals could not expect to achieve such good results. The average false negative rate for malignancy with experienced cytologists is about 5%–10%, and the average reported rate overall is about 10%–15% (literature range, 0%–50%). Follicular carcinoma is more difficult to diagnose than papillary carcinoma. The average false positive rate for experienced cytologists is about 2%–4%, and the average reported rate overall is about 5% (range, 0%–14%). Most pathologists without special interest or extensive experience in fine-needle aspiration cytology are better able to interpret needle tissue biopsy material than thyroid aspiration cytology, because thyroid cytology takes special training and experience. However, well-differentiated follicular carcinoma is difficult to diagnose on needle biopsy as well as on aspiration. Needle biopsy is also useful to diagnose thyroiditis.

Thermography and B-mode ultrasound have been used to help evaluate thyroid nodules for malignancy. Results of thermography to date have been rather disappointing. Ultrasound has been used to differentiate cystic thyroid lesions from solid ones. About 15%–20% of thyroid nodules that fail to concentrate radioactive iodine are cystic. Typical completely cystic lesions are rarely malignant (about 2%; literature range, 0%–14%). Ultrasound accuracy in differentiating pure cystic lesions from solid or mixed cystic-solid lesions is usually quoted as about 95% (80%–100%). The procedure in many clinics is to perform aspiration with cytology on ultrasonically pure cysts.

Medullary carcinoma of the thyroid. Medullary carcinoma constitutes 5% (range, 2%–10%) of thyroid carcinomas. It is derived from certain stromal cells known as “C-cells.” The tumor has an intermediate degree of malignancy. It may occur sporadically or in a hereditary form. The sporadic form comprises 80%–90% of cases and is usually unilateral. The familial variety is transmitted as an autosomal dominant trait, is usually present in both thyroid lobes, and is frequently associated with other neoplasms (phenochromocytoma, mucosal neuromas) as part of MEN II (Sipple Syndrome,Table 33-13) or MEN III. This also includes some degree of association with other endocrine abnormalities, such as parathyroid adenoma and Cushing’s syndrome. The tumor may have a variety of histologic patterns, but the classic form is solid nests of cells that are separated by a stroma containing amyloid. These tumors have aroused great interest, since most secrete abnormal amounts of the hormone calcitonin (thyrocalcitonin). Calcitonin has a calcium-lowering action derived from inhibition of bone resorption; therefore, calcitonin acts as an antagonist to parathyroid hormone. Thyroid C cells produce calcitonin as a normal reaction to the stimulus of hypercalcemia. About 70%–75% of medullary carcinomas produce elevated levels of serum calcitonin; this includes most sporadic (nonfamilial) cases. About 25%–30% of familial medullary carcinoma (MEN type III or IIB) have normal basal calcitonin levels. In patients with normal basal calcitonin levels, elevated calcitonin values can be induced by stimulation with calcium infusion or pentagastrin. Glucagon also stimulates calcitonin secretion but not as effectively. A few medullary carcinomas are reported to secrete serotonin or prostaglandins. About 30% of patients experience diarrhea. Besides medullary thyroid carcinoma, calcitonin secretion has been reported in as many as 60% of patients with bronchogenic carcinoma (small cell and adenocarcinoma tumor types).


Table 33-13 Multiple endocrine neoplasias