Treatment
(Note-The American Thyroid Association has recently issued (2006) useful Guidelines for Management of Differentiated Thyroid cancer)
<www.liebertonline.com/doi/pdf/10.1089/thy.2006.16.ft-1?cookieSet=1>

Treatment related to Stage of disease

In differentiated cancers contemporary medical and surgical practice depends mainly on the clinical stage of the disease rather than on the exact histologic status. The anatomic stage can be defined precisely by the UICC TNM system, version 6 (see Table 18- 4) (370, 371) using information from the clinical and pathologic examinations. The classification can also be conveniently reduced to four categories that have prognostic significance (Tables 18-4 and 18-5) and clear therapeutic relevance. It makes only a small difference in the initial surgical therapy whether a small focus of cancer in the gland is papillary or follicular. Even with 131I therapy it makes little difference, because the therapeutic regimen depends on the function of the metastasis, not the structure of the primary lesion. Although follicular tumors are typically functional, metastases do not always follow the histologic or behavioral pattern of the primary tumor, and one explores the possibility of 131I therapy in most patients, regardless of the pathologic classification.

After the physical examination, neck ultrasound, aspiration cytology, and possibly x-ray examination of the chest and 131I thyroid scintiscan, thyroid tumors can be separated tentatively into four categories (which are further defined at operation): intrathyroid mass only (Clinical Stage I): intrathyroid mass plus enlarged, movable cervical lymph nodes (Clinical Stage II); fixed thyroid mass or fixed metastases in the neck (Clinical Stage III); and metastatic malignancy outside the neck (Clinical Stage IV). (Fig 18-12, Fig 18-14, below)

Which operative procedure is indicated?

The possible surgical approaches range from a simple removal of the nodule to total thyroidectomy with bilateral radical neck dissection. If frozen section and operative examinations disclose no evidence of malignancy, only lobectomy is necessary for the benign lesion. It must be noted that frozen section carries a significant rate of false negative diagnosis, compared to final histology. For this reason, some authors prefer to do total (or near-total) thyroidectomy whenever FNAC is suspicious for thyroid cancer, without performing frozen section. Miller et al examined the use of routine intra-operative pathological diagnosis and concluded that it was specific, reduced costs, and limited the number of completion thyroidectomies necessary in patients with cellular follicular lesions. (Miller MC, Rubin CJ, Cunnane M, Bibbo M, Miller JL, Keane WM, Pribitkin EA. Intraoperative pathologic examination: cost effectiveness and clinical value in patients with cytologic diagnosis of cellular follicular thyroid lesion). Thyroid. 2007 Jun;17(6):557-65)On-the-spot examination should be followed by a careful study of multiple areas on permanent sections, since in many follicular lesions the diagnosis of malignancy can be made only from paraffin sections. Among patients with histologic or gross evidence of cancer within the gland, some will have cervical lymph node involvement and others will have no obvious spread. The surgeon has the option of removing the thyroid totally or in part, and performing a local or more radical excision of the cervical nodes. Prophylactic neck dissection is not done. Some patients will be found with cancer in the thyroid gland or in nodes only after examination of permanent histologic sections. One must then decide if a second surgical procedure is to be performed or an attempt made to treat the tumor with 131I.

Table 18-5. Comparison of Two Clinical Staging Systems

Several other staging systems have been developed. While the Clinical Class system is based simply on the extent of disease, the other systems are designed to predict outcome. The TNM Staging system is noted above. The EORTC classification proposed by the European Thyroid Association is based on age, sex, histology, invasion , and metastases (371a). The Dames classification includes data on age, extent and size of primary, distant metastases, and DNA ploidy(371b). MACIS  includes data on age, invasion, metastases, size, and completeness of surgery(371c). All of the systems appear to be effective in categorizing patients into largely similar low and high risk groups.

An Historical review of approaches to Treatment of Stage I Differentiated Tumors (Intrathyroidal Disease Only)

Clinical Stage I differentiated cancers over 1 cm in size comprise half of total cases, occur in younger individuals, and have an excellent prognosis. After surgery (and often 131I ablation), the survival was in our experience 95 - 98% at 15 years and about 95% at 30 years. Because of this excellent prognosis for treated patients, a controversy has continued during 4 decades over the appropriate extent of surgery, and necessity and value of post-operative 131I ablation (372-387).

Lobectomy--At one extreme may be cited the position of Crile (388, 389) and Cady (390), who suggested that papillary carcinoma should be treated by excision of only the primary tumor (and grossly involved nodes), and afterward by administration of thyroid hormone. Cady doubts that thyroxine is needed and does not advise 131I use. Among 307 patients with papillary cancer treated by Cady, 95% survived for 5 years or more. Only two patients died of their cancers. Buckwalter and Thomas (391), and others suggested that lobectomy combined with dissection of nodes in the tracheosophageal groove is an adequate procedure for a Stage I differentiated cancer without evidence of multicentric involvement. In a series of 322 patients, total lobectomy (or total thyroidectomy) was a more effective procedure than subtotal lobectomy for Stage I well- differentiated cancer (392). Interestingly, there were no cases of multicentricity in Stage I cancer in Buckwalter's series. Shands and Gatling (393), in contrast, found an 11% contralateral recurrence rate after lobectomy, and one-half of these patients died. Tollefsen et al (394) found a 4.6% recurrence rate after lobectomy for papillary cancer, and 7 of 17 patients with recurrence in the remaining lobe died of the disease (395).

More Extensive Surgery--A more aggressive approach for Stage I cancer is an ipsilateral total lobectomy and a contralateral intracapsular subtotal lobectomy with resection of nodes in the homolateral tracheoesophageal groove (396). This operation minimizes damage to the parathyroids and to the recurrent laryngeal nerve on the contralateral side. Black et al. found an 84-90% 10-year survival for Stage I papillary carcinoma and preferred this procedure to unilateral lobectomy because of the incidence of tumor involvement of the contralateral lobe. Rickey and Howard (399) and Buckwalter and Thomas (397) made no distinction between Stage I papillary and follicular cancers, but others treat Stage I follicular carcinoma more radically with total thyroidectomy (398) (v.i.). Prophylactic neck dissection is not recommended.

“Near-total" or total thyroidectomy
"Near-total" or total thyroidectomy is advocated by those who are impressed by the incidence of multicentricity in Stage I cancer, and that 2-5% of these young patients ultimately die of their disease. "Near- total" thyroidectomy refers to a procedure which intentionally leaves small portions of thyroid tissue near parathyroid glands or at the entry of the recurrent nerve into the larynx, and is associated with a marked reduction in possibility of hypoparathyroidism and nerve damage. It is frequently used with intended 131I ablation of residual thyroid tissue. (Fig.18-19 and Fig.18-20)

Multicentric involvement is reported to range from 25 to 90%. The wide variation of multicentricity (or intraglandular dissemination) can be explained in part by the finding that the incidence of multicentricity is doubled if one does whole gland histologic sections. There is little or no relationship between the size of a solitary nodule and the incidence of intraglandular dissemination, but an increasing degree of histologic malignancy is associated with the frequency of dissemination. Surprisingly, the incidence of pathologically detectable multicentricity always greatly exceeds that of clinical recurrence of tumor if the opposite lobe is left intact (400). Whether this difference in incidence is a matter of time or of natural biologic defenses against tumor growth, is uncertain. However, it is clear that multicentricity is not the only factor associated with an improved prognosis after more extensive surgery (and 131I ablation), since in numerous studies multicentricity per se has been shown to have no clear effect on deaths or recurrences.

Mazzaferri et al., in their review of 576 cases of papillary carcinoma, found that total thyroidectomy significantly reduced the incidence of recurrences, and recurrences will presumably be correlated with deaths from disease (401). Samaan et al (402) also supported this procedure. Hay et al recently evaluated the efficacy of different surgical approaches to treatment of patients with low risk papillary carcinoma at the Mayo Clinic and concluded that more extensive surgery was not associated with lower case specific mortality rates, but was associated with a lower risk of local regional recurrence. Their data supports the use of bilateral resection as the preferable initial surgical approach (403).

Total thyroidectomy carries an increased risk of hypoparathyroidism, recurrent nerve damage, and the necessity for tracheostomy (404). Accidental unilateral nerve damage may reach 5%, but fortunately bilateral injury is rare (405). All surgeons attempt to preserve those parathyroid glands that can be observed and spared, and an attempt is often made to transplant resected glands into the sternocleidomastoid muscles. Reports range from a 1 to a 25% incidence of hypoparathyroidism after total thyroidectomy (406, 407).

An alternative procedure for those who wish to have total thyroid removal, but who fear total thyroidectomy, is ipsilateral lobectomy and contralateral subtotal lobectomy followed by a dose of RAI to destroy the remaining remnant. The occurrence of hypoparathyroidism is close to zero after this procedure.

Radical excision--The most radical position advised in past years for the treatment of Stage I differentiated cancer was total thyroidectomy combined with homolateral prophylactic radical neck dissection (408-410). Meissner and co-workers (411) found that when radical neck explorations were done, 65% of the patients without clinically apparent metastases were found by the pathologists to have lymph node involvement. Among 282 patients seen at Memorial Hospital in New York City, 46% with presumed Stage I disease were found in fact to have lymph node involvement in specimens obtained from homolateral radical neck dissection. Of 194 patients undergoing homolateral neck dissections, 177 had malignant nodes, as did 20 of 22 patients with contralateral neck dissection. Of 63 patients who did not have radical neck dissection, 21% later had further evidence of cancer, with a 33% mortality. It is not easy to reconcile such a high incidence of inapparent metastases with the low mortality among patients who did not have neck dissections, unless one assumes that the metastatic tumor in most instances failed to become clinically evident over a period of 10-20 years. Opinion has swung strongly away from prophylactic dissections. It is obviously impossible to predict which patients will derive benefit from this procedure. Nearly all surgeons now avoid enbloc dissections since such a resection is not required for long-term survival in thyroid carcinoma, and the procedure, which is mutilating, is especially undesirable in the young women who most frequently have such tumors.

Several groups have recently established criteria for low risk patients, including patients under 45, lack of invasion or metastasis, tumor size under 3 cm, and (in some analyses) low tumor grade on histologic criteria. These criteria categorize low and high risk patients very effectively (412). Problems with such risk stratification is that the criteria are not available at the moment of surgery, and they fail to predict the few "low risk" patients (2.5-5%) who may behave as do "high-risk" patients, and might benefit from more extensive surgery.

Analysis of data on 269 papillary cancer patients followed at the University of Chicago for on average 12 years indicated that near-total thyroidectomy, or lobectomy plus contralateral thyroidectomy, compared to less extensive procedures, was associated with a statistically significant reduction in both recurrences and deaths in patients with Stage I and II cancers over 1 cm in size.

A Generally accepted Contemporary Approach to Surgery
A current approach to surgery, compatible with analyses of surgical treatment to disease outcomes in numerous recent studies, is outlined in Table 18-6. In unifocal papillary or follicular tumors under 1 cm detected at surgery, lobectomy plus contralateral subtotal thyroidectomy is done. This must be a sufficient procedure (some would say too aggressive) for these patients, who have a death rate (after therapy) not significantly different from that of normal persons over 10-20 years. Probably this procedure can be done regardless of age. This operation is also acceptable in a previously irradiated gland if the residual thyroid tissue appears to be entirely normal. If such minimal tumors are not recognized at surgery, but are found incidentally at pathologic examination of the tissue (e.g., after operation for Graves' disease), reoperation is not required.

In papillary or follicular tumors more than 1 cm, with multicentricity or with a history of neck irradiation associated with multiple abnormalities in the gland, we prefer a near-total thyroidectomy done by a surgeon with expertise in the field. This operation is an extracapsular thyroidectomy, but with the intention of and willingness to leave small amounts of tissue around the parathyroids and recurrent nerves, as necessary, to assure their viability.

Table 18-6. Suggested Surgical Procedures in Thyroid Cancer

* STT = Subtotal thyroidectomy ** NTT = Near-total thyroidectomy *** MND = Modified neck dissection

Stage II Disease

In every procedure the tissue in the tracheoesophageal groove and the supraclavicular fossae should be palpated to the extent possible, and nodes in the homolateral groove should be biopsied. If nodes in this area are malignant, the supraclavicular area should be further explored; if the nodes are malignant, a modified neck dissection should be performed (413-415). We do not favor enbloc dissection. The dissection should include all lymph node-bearing areas in the anterior neck as determined by inspection at surgery. Dissection of the upper mediastinum should be performed if malignant nodes are found, but rarely is a sternum-splitting operation justified. A more formal "block dissection" of involved nodes may be desirable in follicular cancer, since this tumor is more locally invasive. Most physicians who have considerable experience with the problem advise an ipsilateral "modified" dissection of node-bearing tissue, retaining the accessory nerve, jugular vein, and sternocleidomastoid muscle. Although some surgeons advised in the past a radical neck dissection (416, 417), it is obviously a much more mutilating procedure and there is no evidence that it offers a better prognosis. Those who favor a modified neck dissection cite the fact that the disease can usually be clinically cured by this procedure, and that, since the disease usually remains localized to the nodes for a long period of time, recurrent nodes may be removed at a second operation later on if necessary. Rickey and Howard found a 100% 10-year survival of papillary and follicular cancer patients after modified neck dissection (418).
During follow-up recurrent  (or residual) nodes are frequently found in the neck by palpation or US. Localization of non-palpable nodes during re-operation can be difficult. A hook needle implanted under US guidance can help localization and subsequent  removal of non-palpable metastatic nodes (418a).

Management of Stage III Disease

There is less controversy about the extent of required surgery in more advanced tumors. Patients with invasive papillary or follicular tumors should have as extensive a dissection as possible without mutilation. It appears that in young patients residual tumor can be left along one recurrent nerve, for example, to be ablated by 131I therapy, without compromising the patient's life expectancy. Thus, in 97 patients with an incomplete procedure that left tumor in the neck, 83% had a 10- year survival (419). It becomes obvious that survival cannot be directly correlated with the extent of surgery, and a certain degree of conservatism is warranted. In patients over age 45, since the papillary or follicular tumors are clearly more aggressive, more extensive resection including portions of the trachea may be required, and the sternocleidomastoid muscle, jugular vein, and accessory nerve should be removed if this procedure allows apparently complete tumor resection.

Invasion and sometimes obstruction of the trachea, and invasion of the esophagus, are two of the most feared complications of thyroid malignancy and are especially associated with papillary carcinoma in older patients, and with follicular and Hurthle tumors. Nishida et al treated patients with limited invasion by resection without removal of the trachea, and those with deep invasion into the trachea by resection of the invaded trachea. Ozaki et al used circumferential sleeve resection and end-to-end anastomosis to manage patients with thyroid carcinoma invading the trachea. Three to nine tracheal rings were resected. Longitudinal spread of disease on the mucosal side never exceeded that apparent on the adventitial side. Resection was planned to be one tracheal ring proximal and distal to the invasion margin. Release of the trachea by blunt dissection proximal to the larynx and distal to the tracheal carina was fundamental. Sugenoya et al described the use of delayed closure of tracheal cutaneous fistulas developed during surgery by using transplanted auricular cartilage as a free graft to supply tracheal wall skeletal support for the permanent closure (420-422). Tumor control can also be achieved by a variety of other methods including intraoperative radiation therapy, endotracheal hyperthermia, and Irridium-192-brachytherapy, as described by Wolf et al (423). Considering the severe problem of tracheal stenosis, these methods at least offer hope for improvement in the quality of life.

When significant residual tumor is left in the neck, especially in patients beyond age 45-50, external radiotherapy is advised by some authors to control local disease.(v.i.) This practice may be particularly useful when residual tumor does not take up radioiodine, as often is the case.

Management of Stage IV Disease

Finally, one must consider the group of patients with metastases outside the confines of the neck. Those with a thyroid mass and solitary lung or bone metastases should undergo thyroidectomy and excision of the metastases if the latter are accessible. Prolonged survival may be observed. If there are multiple metastases, RAI ablation is attempted in addition to thyroidectomy. Unfortunately, the prognosis in Stage IV disease must be guarded. Patients with 131I uptake in the lungs but no detectable nodules on X-ray or CAT scan tend to do well with 131I therapy (424). However, patients with other soft tissue or bone lesions, especially if multiple, have up to 75% mortality in ten years (425).

It may also be possible and profitable to remove distant solitary metastases. The authors have, for example, removed a solitary metastasis in the skull of a woman who for 11 years after operation was entirely free of evidence of disease. Removal of brain metastasis appears to prolong survival as well.

A problem arises in patients with widely metastatic disease from an unknown primary site that is histologically and clinically compatible with thyroid cancer. (Fig.18-15) If the diagnosis of thyroid cancer can be established, palliation or cure becomes occasionally possible, and accordingly one should take further steps to establish a definitive diagnosis. These steps include immunostaining of the tumor tissue, measurement of TG in the serum or in the fluid aspirated from any accessible lesion, sometimes ultrasound or CAT scans of the thyroid, a 131I scan of the thyroid and metastases, and possibly thyroidectomy. If the scan is abnormal, it can be assumed that the metastsis is probably thyroid cancer. A thyroidectomy is then done to establish a histologic diagnosis. With the thyroid removed, thyroid metastases are more likely to show uptake on scan, and if RAI treatment is contemplated, tumor uptake of the isotope is enhanced. If upon thyroidectomy the diagnosis of thyroid cancer is made, a therapeutic dose of 131-I on TSH stimulation should be attempted and a post-therapy scan should be performed to ascertain whether the lesions take up radioiodine.

After operation all patients are maintained indefinitely on TSH-suppressive thyroid hormone therapy except when undergoing tests. Individuals with current active cancer (other than medullary or lymphoma) should have TSH suppressed to as near zero as is acceptable to the patient. Patients who had cancer previously, but are currently believed free of disease, should have their replacement lowered to provide a TSH of 0.1-0.4 µU/ml, and ultimately as safety is assured, to the normal range. With the exception of persons under age 21 with no history of irradiation and unifocal papillary or follicular tumors less than 1 cm in size, all patients undergo postoperative 131I ablation and 131I therapy, if appropriate, as described below.

Some centers use 131-I ablation in any patient with differentiated thyroid cancer, regardless of age and tumor size. Patients with Stage III papillary or follicular tumors over age 55 also may receive external irradiation therapy, as may patients of any age with Stage__III medullary cancer. All patients with undifferentiated neoplasms or lymphomas, or with nonresectable neck recurrence of papillary or follicular cancer, also receive irradiation, or chemotherapy, as described below.

Radiation-Associated Cancers

Radiation-associated cancers have a high degree of multi-centricity (54% in our series, more than double that in a control group of non-irradiated patients), are usually papillary or follicular tumors, tend to be the same size as cancers in nonirradiated persons, are often associated with other irradiation-induced lesions such as adenomas and fibrosis, and have at least as bad a prognosis as radiation-unassociated cancers. We accept lobectomy and contralateral subtotal thyroidectomy if there is only one lesion under 1 cm in size and the rest of the gland is normal. If there is multicentricity, a tumor more than 1 cm, adenopathy, or multiple abnormalities in the gland (the usual case), near-total thyroidectomy is performed.

Childhood Thyroid Cancer

Some special features of thyroid cancer occurring in children deserve comment. It is, of course, an uncommon disease. The tumors are usually papillary or mixed histologically, and tend to grow slowly, with a high frequency (50 - 80%) of neck metastases, but with a relatively favorable prognosis. Very young patients (under age 12) often have relatively aggressive disease. The association with x-ray exposure has already been discussed. As in adults, the incidence in girls is double that in boys. Multicentricity of tumors is found in 30 - 80%. Metastases to lungs are common (perhaps 20%), but tumor is rarely found in the bones. Lung metastases usually accumulate 131I and can often be eradicated with this isotope, particularly those not visible with X-rays.

As with adult tumors there is no universally accepted surgical approach, but it is certain that sentiment has swung away from prophylactic and radical neck dissections to a more conservative position (426-428). The operations employed are as described above, and near-total thyroidectomy, done by an experienced surgeon, is favored. Thyroid remnants are destroyed with 131I in patients with multicentric lesions and in all Clinical Stage II, III, and IV lesions. Detection of metastases is attempted by 131-I scanning, as described elsewhere in this chapter. Most childhood metastatic thyroid cancers are found to accumulate sufficient 131I to allow useful and sometimes curative therapy, often with doses of 75-150 mCi. Presumably children are more likely to suffer side-effects of 131-I therapy, as described below. Reproductive potential is diminished by large doses of 131-I (429), but an increased incidence of birth defects has to date not been encountered among the relatively few progeny studied (430-432). Thyroid hormone is given to suppress TSH to the 0.1uU/ml area in patients who have known residual or probable residual disease, even though this is known to cause some loss of bone mineral. Although the 10-year survival is from 90 to 95%, long term follow-up demonstrates an eightfold greater than normal mortality (433) and emphasizes the need for comprehensive therapy and long term follow-up.