Abstract
The guidelines begin with a discussion of the apparent increasing incidence of thyroid carcinoma. Indeed, incidence figures for new cases of thyroid cancer in the United States continue to rise with estimates for 2009 of 37,200 new cases and 1630 deaths (8). Davies and Welch (9) proposed that the increased incidence simply reflected increased detection of small tumors; for example, due to increased use of ultrasonography. Recent epidemiologic studies published after the taskforce completed its literature review suggest that the incidence of thyroid cancer is indeed increasing and that it is not just earlier detection of small tumors (10 –12).
Thyroid Nodules
The new guidelines include a very useful table that stratifies risk of malignancy on the basis of ultrasonographic characteristics to determine which nodules warrant fine-needle aspiration cytology (FNA). Incidentalomas of <1 cm are largely dismissed because of the low cost–benefit ratio of full evaluation unless there are factors present that are associated with cancer risk. The guideline recommendation fails to provide a clear recommendation for continued follow-up of these nodules by serial ultrasound examinations. For larger nodules, however, it is clearly stated that benign nodules on FNA are to be serially followed by ultrasound rather than just by physical examination, with the frequency of repeated ultrasound decreasing with time and reassurance of stability of size. As before, measurement of serum thyrotropin (TSH) is recommended in the initial evaluation, with performance of a scintiscan when TSH is found to be suppressed. While the European consensus guidelines (3) recommend calcitonin (TCT) measurement, the revised ATA guidelines recommend neither for nor against screening (recommendation I). Inability to be more definitive is based partly on the issue of cost effectiveness and on the nonavailability in the United States of pentagastrin for stimulation tests that have proven useful for the detection of MTC in Europe. Cost effectiveness of TCT screening is likely to depend upon the size of the nodule, the dollar cost of the assay, and the number of times that the assay might have to be repeated in serial follow-up (13). A recent study suggests that false-positive rates can be reduced and most MTC detected by employing an upper limit of normal of 15 ng/L rather than 10 ng/L (14). Measurement of TCT may be useful for diagnosis of MTC as well in washout fluid of aspirates of thyroid nodules or lymph nodes (15,16).
As in the 2006 publication, the detection of a thyroid nodule is considered a clear indication for ultrasonography, a practice that is still to be adopted by many clinicians who may go directly to FNA under palpation and then to surgery when cytology suggestive or indicative of cancer is seen. Preoperative ultrasonography is recommended for the delineation of the contralateral lobe and lymph node metastases and serves to guide the surgical approach. While also recommended in the 2006 guidelines, this practice is lamentably also not yet widely followed. The new guidelines amplify the utility of ultrasonographic characteristics of nodules that may imply malignancy, and teach us that the classic findings of microcalcifications and hypoechoic appearance apply to papillary thyroid cancer but not to follicular tumors. Then, perhaps ignoring this latter admonition, the guidelines do not recommend FNA for nodules of <1 cm diameter unless they are hypoechoic and/or contain microcalcifications, with risk of missing follicular tumors as a result. Presumably, the greater complacency with smaller follicular tumors relates to the rarity of distant metastases from microcarcinomata. Given that skilled ultrasonographers can readily perform FNA on nodules of 0.5–1.0 cm, clinicians may choose to do so early in the evaluation or wait for evidence of nodule growth during serial periodic ultrasound examinations. Arguments against routinely performing FNA on subcentimeter nodules have been presented by Mazzaferri and Sipos (17).
There is a change in the recommendation for the use of molecular markers such as galectin-3, BRAF, etc., a practice that was not recommended in 2006, but since then the task force has been convinced of their potential utility. Indeed, additional studies attesting to the preoperative value of BRAF testing to prospectively guide surgical planning and follow-up continue to appear (18 –20). As in the 2006 guidelines, preoperative measurement of thyroglobulin (Tg) is not recommended.
Questions frequently arise in regard to which nodules in a multinodular goiter should undergo FNA, and the guideline recommendation has not changed since 2006 insofar as performing FNA on any nodule with suspicious finding on ultrasound, or if there are none, to then perform FNA on the largest nodule with the other nodules to be followed by serial ultrasound examinations. The task force again has strongly recommended against levothyroxine therapy of nodules with benign findings on FNA. A new suggestion is that so called “spongiform” and purely cystic nodules on ultrasound do not require FNA, and that cystic nodules may be a candidate for percutaneous ethanol ablation. Basic recommendations for management of nodules discovered in children or during pregnancy have not changed since 2006.
Thyroid Cancer
Total thyroidectomy is the procedure of choice for nodules with FNA findings clearly suggestive of cancer and for tumors that are very large, demonstrate some suspicious findings, or occur with a history of radiation exposure (the latter being a new recommendation). However, the task force accepts the option of only a lobectomy for patients who request the lesser procedure and who have an indeterminate nodule on cytology as well as for patients with a single microcarcinoma of <1 cm with no other suspicious history or findings on imaging. A distinction has now been drawn between “prophylactic” and “therapeutic” neck dissections. A recommendation for routine central neck compartment (Level VI) node dissection has been strengthened from “should be considered” in the 2006 guideline to “should accompany total thyroidectomy” in patients with enlarged lateral neck nodes, and “may be performed” in those without involved lateral nodes, thereby bringing the ATA guidelines closer to the European ones (4). The recommendations for completion thyroidectomy have not changed since 2006 as well as the recommendations for staging by the American Joint Committee on Cancer–International Union Against Cancer classification system (21).
A recommendation (R21) for preoperative ultrasonography with FNA of suspicious lymph nodes is also consistent with the European consensus guidelines.
Recommendations in the 2009 guidelines for radioactive iodine ablation remain largely the same as in 2006 but are much less ambiguous with specific criteria tabulated to indicate specific criteria as to when to ablate Stage I tumors in respect to potential benefit based on tumor size, characteristics, etc. One important distinction in regard to low-risk patients deserves mention. Previously, ablation was recommended for patients with multifocal microcarcinomata, presumably because of the possibility that another microscopic tumor could be present in thyroid remnant tissue. Although a recent study indicates recurrence is more frequent in patients with multifocal microcarcinomata than in unifocal disease, radioiodine therapy was not associated with fewer recurrences (22). Consequently, with no benefit of radioiodine shown in patients with multifocal tumors, the new guidelines do not recommend routine ablation. Ablation is also not recommended for a single tumor <1 cm unless there are worrisome aspects for higher risk such as lymph node metastases or a histological variant such as tall cell or insular tumors that are associated with higher recurrence rates and worse outcome.
Guidelines for thyroid hormone withdrawal in preparation for ablation are presented as before (Guideline R33) including the statement that levothyroxine (with or without liothyronine) may be resumed on the second or third day after the radioiodine. Given that TSH levels after withdrawal may remain elevated for 1–2 weeks after resumption of thyroid hormone, and because TSH will enhance discharge of bound 131I from thyroid remnant or thyroid cancer thereby potentially mitigating therapeutic effect, the rationale for delaying treatment for 2–3 days is puzzling, for it appears more reasonable to initiate therapy within 12 hours of the radioiodine therapy. Of course, the option of preparation for ablation employing recombinant human TSH (rhTSH; Thyrogen®, Genzyme Corp., Cambridge, MA) was not available in 2006. With approval of Thyrogen for this purpose in December 2007, this approach is now “strongly recommended” by the guidelines, although no schedule or protocol for rhTSH ablation is provided. In addition, no distinction is made between ablation for disease considered Stage I versus ablation in patients with findings suggesting Stage III or IV disease with higher risk (R34). Currently, most centers confine use of rhTSH preparation for ablation to low-risk disease based on the literature extant on the efficacy of rhTSH-mediated ablation in clinical trials that largely included low-risk patients. Use of rhTSH for ablation in Europe is also confined to low-risk patients (4). Moreover, the Thyrogen® label specifically precludes its use for ablation in patients with metastatic disease, which would include Stage II patients under age 45, and Stage IV patients over age 45, and this restriction is not mentioned in the guidelines. Of important significance in respect to ablation mediated by rhTSH is the lower radiation dose than that which occurs by withdrawal as shown by the study by Pilli et al. (23) cited in the guidelines; this has been confirmed by two other recent studies not cited by the task force (24,25).
The European consensus guidelines questioned the utility of pre-ablation diagnostic scans based upon a potential risk of stunning and the belief that the post-therapy scan provided all needed information (3). As pointed out by Van Nostrand et al. (26) and by McDougall (27), use of 123I for scanning obviates the risk of stunning and useful information is often provided by the pretherapy scan. On the basis of “expert opinion,” the new ATA guidelines do recommend pre-ablation diagnostic scans under a number of specified conditions. The new guidelines do not alter the earlier recommendation for 131I dosage of 30–100 mCi for low risk tumors and 100–200 mCi for more aggressive tumors.
The recommendation for a low-iodine diet for 1–2 weeks prior to ablative therapy is unchanged from the 2006 version, along with the suggestion to measure a spot urine iodine to ensure compliance prior to therapy. Unfortunately, this is impractical because urine iodine assays are not performed routinely and the result will not be back by the time of scheduled therapy. Instead, it is preferable to have patients adhere to 3 weeks on the diet and measure a spot urine after 2 weeks thereby leaving adequate time to see the result prior to therapy. An excessive excretion of iodine in the urine should result in postponement of the therapy. Unmentioned in the guidelines is one important caveat that applies to hypothyroid patients undergoing thyroxine withdrawal and low-iodine diet prior to radioiodine therapy. These patients may suffer dangerous hyponatremia due to mistaken avoidance of all salt rather than avoiding only iodized salt and should be advised to maintain normal salt intake (28).
There is an excellent discussion that is largely unchanged from 2006 of the utility of Tg measurements during long-term surveillance and the problem of interfering anti-Tg antibodies. Based upon recent literature, it is stated that a Tg of <0.5 ng/mL with negative antibodies is associated with an approximate 99% rate of cure. It is important to point out that 99% is not 100%, and Stage III or IV patients are more likely to be the exceptions, emphasizing the importance of continued monitoring. Employing more sensitive Tg assays, assurance of a full cure will likely be more certain when a rhTSH-stimulated Tg of <0.1 ng/mL is observed. The new guidelines address how often we need to repeat rhTSH stimulation tests and better define what represents absence of tumor based upon serum Tg on suppression and findings on ultrasonography. Since publication of the 2006 guidelines, many studies have assessed and documented the usefulness of neck ultrasonography in surveillance for recurrent or residual disease and this literature is well summarized in the new guidelines. Indeed, the new guidelines should result in a reduction in the frequency of ultrasound examination compared to recommendations in 2006 as well as a reduction in rhTSH stimulation testing and FNA, and a greater role for 2-deoxy-2[18F]fluoro-
In regard to patients with metastatic disease who are candidates for additional 131I therapy, the 2006 guidelines made no recommended choice between dosage selection by empiric means or by dosimetry. The new guidance is unchanged with one exception, i.e., taking heed of the risk of overdosage with empiric dosage of 200 mCi or greater in an elderly population. This admonition is based on the results of two studies (30,31) that demonstrated risk of excessive radiation from empiric dosage of radioiodine. Although the risk of overdosage is less with empiric doses of 125–175 mCi, it is still significant in an elderly population and the indication for dosimetric dosage determination could have been more strongly stated. Other patients at risk for excessive radiation exposure are children and those with pulmonary metastases. The recommendation of the new guideline (R57, R58), that patients with either pulmonary micro- or macrometastases may be treated with empiric doses of 100–200 mCi falls short of optimal, and dosimetry is clearly preferable in these patients.
It is well known that radioiodine is concentrated in salivary glands and that the risk of sialadenitis, loss of taste, and xerostomia are increased further with either higher or cumulative dosage. Many centers use various sialogogue agents in an attempt to reduce radiation exposure risk and the current guidelines make no recommendation for or against their use. The hesitancy is based in part on one study indicating greater salivary damage when sour lemon candies were administered 1 hour after administration of 131I (32). Based upon perceived flaws in the latter study, Van Nostrand et al. (33) restudied salivary sialography after tracer dosage and concluded that radiation exposure is likely reduced by sialogogues such as lemon candies. A larger series of patients will be required to clarify the issue, as pointed out in the accompanying editorial (34). The section on secondary malignancies after radioiodine has been expanded with recommendations for cancer screening. The recommendation for women to avoid pregnancy for 6–12 months after radioiodine is unchanged as is the recommendation to avoid radioiodine for 6–8 weeks after cessation of lactation. However, the use of dopaminergic drugs to reduce lactation was dropped from a “B” to a “C” recommendation.
The section on management of the Tg-positive and scan-negative patient is somewhat expanded but the difficulty in managing these patients continues. Given the flurry of clinical trials with new chemotherapeutic agents that might target thyroid cancer, a significantly expanded section on promising new agents for targeted therapy may have been expected. It appears that this will have to wait for the next revision. There is, however, a more specific recommendation for clinicians to consider entering patients into clinical trials, as well as suggesting that tyrosine kinase inhibitors could be used off label (R61C). As in the 2006 guidelines, some very useful and important strengths of this revision are the extremely concise algorithms for management and new tables. There are new tables on nodule characteristics on ultrasound and factors influencing decision making for remnant ablation with 131I, and a useful figure depicting the node-bearing surgical compartments and levels of the neck. As before, the format and presentation of information is conducive to ready reference. However, because the pace of discovery of new approaches to the diagnosis and treatment of thyroid cancer is certain to continue (if not accelerate), these guidelines cannot be considered enduring. Clinicians should look forward to periodic refinements and revisions in the decades to come that may ultimately resolve many of the remaining controversies associated with management of both benign and malignant tumors of the thyroid. The 2009 guidelines will be invaluable until the next revision and we should all be extremely grateful to the task force for this extraordinary contribution to the care of our patients with thyroid nodules and thyroid cancer.