ABSTRACT: The incidence of papillary thyroid cancer has been rising steadily over the past 3 decades. Most tumors in this setting are regarded as low risk, but recurrence rates are high, producing controversy about initial therapy. Microcarcinomas smaller than 1 cm are generally best treated with lobectomy alone. Total thyroidectomy should be performed for tumors 1 cm or larger or for tumors that have metastasized. Prophylactic central and lateral neck lymph node compartment dissection uncovers unsuspected metastases in about half the patients, which may alleviate the need for postsurgical radioiodine therapy but can be associated with surgical complications. Radioiodine may diminish tumor recurrence but is complicated by injury to nonthyroidal tissues and by dose-related nonthyroidal cancers that occur in a small number of patients. Tumors that are metastatic, invasive, or multifocal or have aggressive histologic features should be treated with radioiodine. Total-body irradiation can be significantly reduced by preparing the patient with recombinant human thyrotropin and by using smaller amounts of radioiodine (~30 mCi). The natural history of papillary cancer is such that patients who achieve disease-free status after total thyroidectomy and radioiodine therapy usually achieve normal life expectancy.
Patients with papillary thyroid cancer typically undergo a triad of consecutive initial treatments, comprising surgery, radioiodine therapy, and thyroid hormone suppression of serum thyrotropin (thyroid-stimulating hormone, or TSH). Each modality, albeit effective, is laced with controversy when applied to the initial treatment of low-risk papillary thyroid cancers. The aim of this article is to address the current controversies surrounding the management of low-risk papillary thyroid cancer in light of the current literature and to provide an analysis of contemporary opinions and recommendations for initial therapy.
The Epidemiologic Controversy
| • | What is optimal surgery for a 2-cm papillary thyroid cancer without preoperatively apparent cervical lymph node metastases? |
| • | What is the role, if any, of postoperative radioiodine therapy, and if so treated, should patients undergo thyroid hormone withdrawal or preparation with intramuscular recombinant human TSH? |
| • | What is the role of postoperative thyroid hormone suppression of TSH? |
| • | What is the optimal means of identifying residual thyroid cancer? |
The initial management of thyroid cancer has attracted considerable attention in recent years, mainly because of its relentlessly rising incidence, which increased 2.4-fold from 1973 to 2000.[1] This is almost entirely attributable to a nearly threefold increase in papillary thyroid cancer, which is by far the most common form of the disease, comprising over 80% of all thyroid cancers (throughout this review, percentages are rounded to the nearest integer).[2]
Between 1988 and 2002, when thyroid cancer tumor size was first included in the National Cancer Institute’s Surviellance, Epidemiology and End Results (SEER) database,[1] it became apparent that nearly half the thyroid cancers diagnosed during this period were papillary microcarcinomas 1 cm or smaller, and nearly 90% were papillary cancers 2 cm or smaller.[3] Tumors of this size are widely acknowledged to be associated with low mortality rates, and the overall mortality rates for thyroid cancer during this period remained stable at 0.5 deaths per 100,000 persons in the population.[3]
It was therefore suggested that these trends—combined with the substantial reservoir of occult thyroid cancer in the general population—are a manifestation of the detection of subclinical disease, not an increase in the true occurrence of thyroid cancer.[3] The authors of this study expressed understandable concern that overdiagnosis makes it difficult to identify which patients need treatment. In addition, most of the patients being studied underwent total thyroidectomy, even if the papillary cancer was very small. The authors consequently recommended that a more cautious diagnostic approach (perhaps simply providing follow-up for symptomatic thyroid nodules) is worthwhile, and that papillary cancers smaller than 1 cm could be classified as a normal finding.[3]
Several issues regarding the epidemiology of small thyroid cancers require further explication. The age of patients at the time of thyroid cancer diagnosis ranges from 4 years to 85 years or older, peaking at 40 to 45 years in women and 50 to 55 years in men.[1] The notion that there is a large reservoir of occult nonthreatening thyroid cancers stems from autopsy studies that report a high prevalence of undiagnosed thyroid cancers.[4] Yet it is highly unlikely that data from autopsy studies from around the world accurately reflect the rates of clinical thyroid cancer in the United States population, considering that the occult thyroid cancer prevalence rate in 20 autopsy series ranges from 2% to 36%.[4] The autopsy study[5] with the highest rate of occult thyroid cancers (36%) found that 67% were discovered in nodular goiters as compared with 24% in otherwise normal thyroid glands, and almost half were in individuals 61 years of age or older.[5] In contrast, clinical cases of thyroid cancer are found with equal frequency in multinodular goiters and single thyroid nodules; furthermore, autopsy studies in younger persons find considerably fewer (2%) occult thyroid cancers.[4]
Also, the gender distribution of thyroid cancers at autopsy is nearly equal,[4] whereas in clinical studies the rate in women is threefold that of men, suggesting ascertainment bias in the autopsy studies. Small thyroid cancers incidentally found at surgery for benign thyroid disease is another matter. The rate of unanticipated microcarcinomas in surgical thyroid specimens is variable, ranging from approximately 2% to 22% in 11 series.[4] This is stronger evidence for a reservoir of very small papillary thyroid cancers in the general population that might explain some of the increase in papillary microcarcinomas, but the eventual outcome of such small tumors cannot be ascertained with certainty.
While overall mortality rates from thyroid cancer have been stable for the past 3 decades, this does not fully describe the outcome in subsets of patients.[6] From 1995 through 2000, the 5-year relative survival rates for thyroid cancer were 97.3% in women and 93.5% in men, which translates into a thyroid cancer mortality rate in men that is more than twice that in women (P < .05).[1] From 1992 through 2001, the annual percent change in death rates decreased 0.03% in women and increased 2.3% in men (P < .05).[1] This was the largest increase in cancer mortality of all malignancies in men in the United States,[1] which can be largely attributed to delayed diagnosis and significantly more advanced tumor stage at the time of diagnosis in men.[1,6] Still, 5-year mortality rates in those age 45 years or younger are very low (approximately 1%).[1]
What are the clinical implications of these epidemiologic findings?
Mortality rates for thyroid microcarcinomas and tumors smaller than 2 cm are very low. Identifying patients who are likely to remain disease-free without therapeutic intervention, however, is another matter. A large study of watchful waiting found that careful observation without surgery might be acceptable for incidentally discovered low-risk papillary microcarcinomas but that careful and systematic surgery should be performed for invasive tumor or lymph node metastases.[7] A study[8] of over 2,000 patients with microcarcinoma followed for up to 40 years found that younger patients with smaller tumors have a better prognosis than older patients with larger tumors (6 to 10 mm). In all, 40% of patients older than 55 years had a recurrence after 30 years of follow-up, as compared with less than 10% of younger patients.
To avoid this problem of microcarcinomas, the American Thyroid Association (ATA) evidence-based guidelines[9] recommend performing fine-needle aspiration biopsy (FNAB) for tumors 1 cm or larger unless the patient is at high risk for thyroid cancer (such as those with a history of head and neck radiation), in which case tumors > 5 to 9 mm can be considered for FNAB. Even then, the natural history of the smaller tumors in an individual patient is uncertain.
Tumor Staging Systems
TNM Classification System for Differentiated Thyroid Cancer
Problems inherent to thyroid cancer staging systems complicate initial management decisions. According to the 6th edition of the American Joint Commission on Cancer (AJCC) tumor-node-metastasis (TNM) staging system,[10] patients under 45 years old with papillary thyroid cancer are classified as stage I (low risk) if there are no distant metastases, regardless of the tumor size, invasion, or locoregional metastases (Table1); stage II is defined by distant metastases. Patients age 45 years or older are classified as stage I if the primary tumor is smaller than 1 cm and is limited to the thyroid without metastases, and are classified as stage II if the primary tumor is larger than 1 cm but not larger than 4 cm and limited to the thyroid. The criteria for stages III and IV in older patients are shown in Table 1.
A number of other prognostic staging systems also have been used to predict the outcome of papillary cancer[11]; however, none account for more than a small portion of the uncertainty in predicting thyroid cancer outcome, and there is no statistically significant superiority of any system over that of the AJCC TNM classification.[11,12] There also are statistically significant differences between papillary and follicular thyroid cancer assessed in different tumor staging systems, with each possessing its own set of independent prognostic factors for cause-specific survival,[13] underscoring the inadequacy of these staging systems in individual patients. The ATA guidelines[9] suggest that TNM staging should be used for all reports of the treatment and outcome of patients with thyroid cancer because the classification is universally available and widely accepted.
Current risk stratification systems rely almost exclusively on clinical, pathologic, and imaging data obtained during the initial evaluation and therapy. Still, none adequately incorporate various histologic subtypes of thyroid cancer, or other variables that might alter prognosis such as mitoses, areas of tumor necrosis, or molecular characteristics of the primary tumor. Moreover, the current staging systems are static representations of the patient at the time of presentation and are not easy to modify as new data become available during follow-up. It is important to recognize that prognosis shifts over time, depending on the inherent biologic properties of the tumor and its response to therapy, which provide much more practical information than does static postoperative assessment. How should clinical risk be assessed?
Thyroid Cancer Recurrence and Mortality
The primary problem with planning therapy for papillary cancer according to staging systems is that virtually all patients under age 45 years with papillary thyroid cancer are identified as being at low risk only on the basis of cancer-induced mortality. Although survival rates are typically favorable in this group, the disease in young patients is often marked by one or more recurrences before the patient can be rendered free of disease. Current staging systems are insensitive to tumor recurrence, a major source of morbidity that seriously impacts the patient’s quality of life. There is little doubt that persistent or recurrent posttreatment locoregional recurrences are a major part of the management of papillary cancer (Figure 1).
The ATA guidelines[9] define low-risk patients as having none of the following tumor characteristics: (1) local or distant metastases; (2) residual macroscopic tumor; (3) tumor invasion of locoregional tissues or structures; (4) aggressive histology such as tall cell, insular, columnar cell carcinoma, or vascular invasion; and (5) radioiodine (131I) uptake outside the thyroid bed on the first posttreatment whole body scan. This is a much more flexible definition of outcome, which is more closely related to the actual evolution of this disease.
