Thyroid cancer accounts for more than 90% of all endocrine malignancies and for more than 50% of all deaths from endocrine cancer. The incidence of thyroid cancer has been increasing since the 1930s, most likely because of the practice of therapeutic irradiation in childhood. The abrupt rise in incidence around 1974 was probably due to the institution at that time of vigorous examination of individuals who had had prior neck irradiation .
American Cancer Society statistics for the past 25 years suggest that the incidence of thyroid cancer has risen slowly since the mid 1970s. Disease incidence increased from 7,800 cases per year in 1974 to 12,500 cases per year in 1992, with more than two-thirds of new cases occurring in women. However, mortality from thyroid cancer has changed little during this period (Figure 1).
Most patients have localized or regional disease at the time of diagnosis, and the 5-year survival rate has exceeded 80% since 1960 and 90% since 1970. Therefore, thyroid cancer remains a relatively modest oncologic concern, accounting for less than 2% of all cancers since 1974 and for less than 0.5% of all deaths from cancer (Figure 2).
Among patients with thyroid cancer, approximately 60% to 70% have differentiated, papillary thyroid cancer. This article presents a broad clinical overview of papillary thyroid cancer and emphasizes disease features that modify prognosis and affect clinical management.
The risk of radiation-induced thyroid cancer has been well documented; data on other causes of thyroid cancer are sparse. Prolonged thyroid-stimulating hormone (TSH) elevation has been implicated as a potential risk factor, but patients with primary hypothyroidism do not appear to have an increased risk of papillary thyroid cancer. Thyroid-stimulatory immunoglobulins present in patients with Graves' disease have also been implicated. Associations with Hashimoto's thyroiditis, Graves' disease, and multinodular goiter have been reported. However, causative relationships between these diseases and papillary thyroid cancer remain poorly documented.
In a recent epidemiologic review of thyroid cancer in the United States , important differences in incidence emerged among ethnic groups. Thyroid cancer was less common in Puerto Rican Hispanics and blacks than in white persons, but New Mexican Hispanics and Asian-Americans had higher rates. Genetic susceptibility is unlikely to account for all the differences. For example, Asians living in the United States are at higher risk for thyroid cancer than Asians living in their native countries.
Although papillary thyroid cancer is generally considered a sporadic disease, it is important to keep in mind that disease clusters have been reported in kindreds with familial polyposis and Gardner's syndrome . Reports of familial disease in the absence of either a hereditary syndrome or neck irradiation increase the likelihood that genetic predisposition plays a role in pathogenesis. In support of this thesis, several recent studies have demonstrated complex structural chromosomal aberrations and cytogenetic abnormalities of chromosome 10q in association with papillary thyroid cancer .
Irradiation in Childhood
During the early decades of the 20th century, low-dose irradiation (usually < 2.0 Gy) to the head and neck was often used for the treatment of benign childhood conditions, such as thymic enlargement, tonsillitis, otitis, hemangiomas, ringworm, and acne. This medical practice has clearly emerged as an important risk factor for the development of papillary thyroid cancer and is considered by many to be responsible for increasing its incidence.
The excess relative risk of thyroid cancer appears to depend on total radiation dose, fractionation schedule, and patient age at the time of irradiation . To date, medical use of radioactive iodine-131 has not been shown to confer a significant risk of papillary thyroid cancer (Table 1). A persistent concern has been the carcinogenic potential of radioactive fallout. However, careful review of available studies reveals methodologic inconsistencies that may have exaggerated risk estimates .
The time between the irradiation and diagnosis of the thyroid tumor averages 10 years but may be more than 30 years. In a recent study, Schneider et al  demonstrated that radiation-induced thyroid neoplasms are rare in the first 5 to 10 years after treatment and cluster between 20 and 40 years post-therapy. Women have a 40% higher rate of radiation-induced thyroid cancer than men, but the slopes of dose-response curves, which reach maximal rates 25 to 29 years after radiation exposure, appear similar for the two genders.
Increased awareness of thyroid radiosensitivity, especially in children, has resulted in the virtual elimination of irradiation for benign conditions and is generating a reassessment of medical radiation use. Thus, radiation-related thyroid cancer is expected to subside as a public health concern in the coming years.
Radiotherapy for Hodgkin's and Non-Hodgkin's Lymphomas
The efficacy of radiotherapy in controlling Hodgkin's and non-Hodgkin's lymphomas has been instrumental in achieving high cure rates. Consequently, radiotherapy is likely to continue as an important component of the therapeutic armamentarium for these cancers. Because many young cancer survivors can expect a near-normal lifespan after neck irradiation, it is important to define their particular risks of thyroid cancer and to design appropriate surveillance and therapeutic strategies.
Radiation doses to the thyroid of more than 200 cGy, especially when given at a young age, have been linked to an increased risk of both benign and malignant thyroid neoplasms . Tucker et al found a 13-fold increase in the risk of thyroid cancer in this group , and Hancock et al noted that benign thyroid abnormalities are even more common . Systematic evaluation of the irradiated thyroid gland reveals significant pathologic  as well as radiologic  abnormalities.
We also have found a heightened risk of thyroid cancer after therapeutic neck (mantle) irradiation for Hodgkin's disease during childhood (Table 2). Among 166 patients irradiated at or before age 16 years and for whom a minimum of 15 years of follow-up is available, 12 have been treated for thyroid cancer to date (unpublished data; patient population identified through a search of the database maintained by the Department of Patient Studies at M.D. Anderson). The presence of papillary thyroid carcinoma in 7% of irradiated Hodgkin's disease survivors exceeds the prevalence of thyroid cancer in the general population, and suggests that prior therapeutic irradiation represents an important risk factor for thyroid cancer.
In our patients, thyroid cancer was found 7 to 19 years after irradiation by investigation of either palpable thyroid nodules or ultrasound abnormalities; all have proven to be papillary thyroid cancers, and all patients remain alive. During the same period and using the same database, only two cases of thyroid cancer were diagnosed among 750 adults treated for Hodgkin's disease, emphasizing the importance of age as a cofactor for the carcinogenic potential of radiation. Fortunately, patients with radiation-related papillary thyroid cancer appear to have the same excellent prognosis as do patients with sporadic disease, even though the former may have more extensive disease at diagnosis .
Because hyperparathyroidism also may occur with increased frequency after neck irradiation , it is particularly important to investigate previously irradiated patients very carefully for parathyroid dysfunction before thyroid surgery, to avoid the potential need for reoperation.
Papillary thyroid cancer is predominantly a sporadic disease. Patients typically present with a dominant thyroid mass or enlarged cervical lymph nodes . Thyroid nodules are relatively common, but less than 15% of clinically detectable solitary thyroid nodules are malignant [14,15].
Histologically, many tumors have both papillary and follicular elements and are referred to as "follicular variants of papillary carcinoma;" such tumors are classified as papillary lesions because their clinical behavior resembles that of pure papillary cancers. In adults, the size of papillary thyroid cancer can vary considerably, from microscopic cancers to large tumors that may invade the thyroid capsule and even infiltrate adjacent structures.
Up to 40% of adults present with cervical lymph node metastases; the figure is much higher in persons younger than 20 years. At the time of presentation, affected lymph nodes are generally on the same side as the primary tumor, but bilateral and mediastinal metastases are encountered in a few patients. Although regional lymph node invasion is fairly common, there is very little tendency toward systemic dissemination.
Sites of distant metastases, in order of decreasing frequency, are the lung, bone, and, rarely, other soft tissues. Older patients are more likely to present with invasive tumors and have a higher risk for distant metastases [16,17].