Prevalence

The high prevalence of micropapillary cancer has been appreciated from autopsy studies done decades prior to the emergence of high-resolution ultrasonography. In the United States, these studies have shown up to a 13% prevalence of micropapillary cancer,⁴ whereas in other parts of the world substantially higher prevalence rates have been noted. For example, in Finland, the prevalence in one study was 36%, leading the authors of that study to conclude that the “smallest forms of occult papillary carcinoma of the thyroid are so common in Finland, that they can be regarded as a normal finding.”⁵ The prevalence of micropapillary carcinoma in pathologic specimens is also highly dependent on how carefully one looks for it. In one Spanish study, the initial prevalence based on grossly visible lesions was 5.3%, but when each thyroid was cut into blocks and carefully examined histologically, the prevalence increased to 22%.⁶ The prevalence of micropapillary carcinoma in some series was independent of age. For example, in Sweden the prevalence was approximately 7% for patients under age 50 or over age 80,⁷ and in Wisconsin in the United States the prevalence was 3% in an autopsy study of young adults.⁸ Micropapillary carcinoma is frequently an incidental finding at the time of thyroid surgery and has been reported in 2% to 24% of surgical specimens.⁹

An Epidemic?

The Surveillance, Epidemiology, and End Results (SEER) database of the National Cancer Institute in the United States provides data that help to define the present and predict the future epidemic of micropapillary cancer.¹⁰ In 2007, the measured prevalence of thyroid cancer (all thyroid cancers, not just micropapillary cancers) was 434,256. Yet if we apply an estimated figure of 6% for the prevalence of micropapillary cancer based on autopsy studies in the United States, the prevalence should be in excess of 18 million individuals. The annual incidence of thyroid cancer in the United States has been increasing, with the 2007 incidence from the SEER database of 11.99 per 100,000, compared to only 4.85 per 100,000 in 1975. In contrast, mortality from thyroid cancer has remained constant. Although there is controversy as to whether the true incidence of thyroid cancer is increasing in the United States,¹¹ it is widely accepted that a major component of the increased thyroid cancer incidence is due to ascertainment bias from improved imaging, allowing us to more readily detect both micropapillary as well as larger thyroid cancers. For example, a Japanese report that utilized ultrasonographic screening followed by fine-needle aspiration biopsy found a 3.5% prevalence of papillary cancer in women over age 30, 75% of which were under 15 mm.¹² If indeed there are 18 million individuals with micropapillary cancer in the United States, we presently have detected less than 2.5% of these cancers. With continued improvement in imaging and aggressive use of ultrasound-guided fine-needle aspiration biopsy, it should not be a surprise that the annual incidence of cancers detected has more than doubled since 1975.

Pathologic data from hospitals support the hypothesis that detection of micropapillary carcinomas accounts for a substantial portion of the increasing incidence of thyroid cancer. For example, at the Queen Elizabeth Hospital in Hong Kong, the percentage of micropapillary carcinomas in operative pathology specimens increased from 5.1% in the period from 1960 to 1980 to 21.7% in the period from 1991 to 2000.¹³ Publications in 2006 from the University of Wisconsin in the United States and at the University of Ferrara in Italy reported that micropapillary cancers represented 43% and 40% of surgically excised thyroid cancers, respectively.^2,3

Therefore, when reviewing data on the natural history of micropapillary cancer, it is important to understand that published series are reporting only on the 2.5% of cancers that have come to clinical attention for one reason or another and that many recent publications study micropapillary cancers that would have gone undetected previously.

Observational Data

Important information regarding the natural history of micropapillary carcinoma has been obtained by the ongoing observational trial of Ito from Japan.¹⁴ Among 1395 patients with biopsy-proved micropapillary carcinoma, after excluding those patients with tumors adjacent to the trachea, posterior tumors that might be adjacent to the recurrent laryngeal nerve, tumors associated with lateral compartment nodes, and those with high-grade histology, 340 patients chose observation over surgical excision. At baseline, 28% of the tumors were multifocal, and 9% had suspicious central nodes; these were not exclusion criteria. During a mean of 74 months (range 18 to 187 months), 9% of tumors grew to greater than 10 mm, 9% grew by more than 3 mm (6% after 5 years of observation, 16% after 10 years of observation), and 2% developed lateral compartment lymph nodes (1.4% after 5 years, 3.4% after 10 years). Ultimately, 109 of the 340 patients opted for surgery for one reason or another (growth was the most common reason). The surgical outcomes were no different for the patients who had surgery at the time of initial diagnosis and for those who had surgery after a period of observation. None of the 109 patients who had surgery after observation had recurrences during an average follow-up of 76 months. In contrast, 32 of the 1055 patients (3%) who had surgery at initial diagnosis had recurrences: 2.5% in cervical nodes, 0.6% in the thyroid bed, and 0.1% in distant organs.

This trial demonstrates that the majority of micropapillary carcinomas did not grow over an average follow-up that exceeded 6 years and was as long as 15 years, and those that did grow, grew slowly. Only a small percentage developed lateral compartment nodes during the period of observation. These data strongly argue that the assessment of the majority of subcentimetric thyroid nodules by fine-needle aspiration biopsy is unnecessary. If such a policy is widely accepted, perhaps the increasing annual incidence of micropapillary thyroid cancer can be reduced.

Clinical Series of Patients with Micropapillary Cancer

There have been many published series of patients with micropapillary cancer from single institutions. The Mayo Clinic recently updated its series, which includes 900 patients with an average follow-up of 17.2 years (range of 6 to 89 years).¹⁵ Twenty-three percent of the tumors were multifocal, 17% bilateral, 2% extrathyroidal, 30% had nodal involvement, and 0.3% had distant metastatic disease. Less than 25% were under 5 mm and over a third were 9 to 10 mm. The 40-year cause specific mortality was 0.7%—all three patients who died presented with lymphadenopathy, one had massive lymphadenopathy, and one had pulmonary metastases upon presentation. Recurrences occurred in 8% of patients, most in cervical nodes, but 1.5% occurred in the thyroid bed. Nodal recurrences occurred in 16% of patients with positive nodes at presentation and only 0.8% of patients without nodes at presentation. Recurrences occurred in 11% of patients with multifocal disease and 4% of patients with unifocal disease.

The Noguchi Thyroid Clinic in Japan also recently updated its series, which included 2070 patients with an average follow-up of 15 years.¹⁶ Recurrences occurred in 3.5% of patients at a mean of 10.3 years. Distant metastases occurred in only 0.2% of patients. Recurrence was more likely in patients with larger tumors (greater than 5 mm), more nodes, and invasion (e.g., into the recurrent laryngeal nerve or esophagus), and less likely in patients with coexistent thyroid autoimmunity.

The series of 281 patients from the Gustave-Roussy Institute in France includes some patients with subcentimetric follicular thyroid cancers but demonstrates similar rates of recurrence: 2.5% in cervical nodes and 1.4% in the thyroid bed.¹⁷

The series of 203 patients from the Queen Elizabeth Hospital in Hong Kong reports a 4.9% rate of nodal recurrence and a 1% rate of local recurrences.¹⁸ Two patients developed pulmonary metastases (1%), and two patients died. The risk of nodal recurrence was increased 6.2-fold when nodes were present at presentation and 5.6-fold when the tumor was multifocal. The researchers did not find higher recurrence rates in tumors greater than 5 mm, but the larger micropapillary cancers were more likely to have extrathyroidal extension. However, comparing micropapillary cancers to macropapillary cancers, there were similar rates of multifocality, but the macropapillary cancers were associated with higher rates of nodal metastases and nodal, local, and distant recurrences.

Recurrence occurred in 4.8% of 293 patients reported from South Korea after a median follow-up of 65 months; cervical nodes at presentation were associated with an increased risk of recurrence.¹⁹ Recurrence occurred in 3.1% of 287 patients from Rome, Italy, and included 2 patients (0.7%) with distant metastases; multifocal disease, extrathyroidal extension, and a higher number of cervical nodes at presentation were risk factors for recurrence.²⁰

Data from several series ^3,15–25 demonstrate multifocality in 20% to 40% of patients, bilateral disease in 10% to 19% of patients, extrathyroidal invasion in 2% to 38% of patients, cervical nodal involvement in 17% to 43% of patients, and distant metastases in 0% to 3% of patients (Table 19-1). In one series of 671 patients from Seoul, Korea, 24% had central nodal involvement and 3.7% had lateral nodal involvement.²⁵

Table 19-1 Presenting Characteristics of Micropapillary Thyroid Carcinoma^3,15–25

^* In one study, 24% had central nodes and 3.7% had lateral nodes.²⁵