Characteristics of Incidentally Discovered Thyroid Cancer

Importance  The incidence of thyroid cancer has been steadily increasing; however, no clear reason for the increase in incidence has been identified.

Objectives  To compare incidentally discovered (ID) thyroid cancer via non–thyroid-related imaging with nonincidentally discovered (NID) thyroid cancer, as well as determine if differences in tumor characteristics and patient presentation in ID thyroid cancer may help elucidate the increasing incidence of this disease.

Design, Setting, and Participants  Retrospective medical record review at an academic tertiary care medical center of 31 patients with ID thyroid cancer and 207 patients with NID thyroid cancer evaluated at our institution during a 12-month period.

Main Outcomes and Measures  Patient demographics, tumor pathology, stage, tumor size, invasion, and metastasis were recorded.

Results  Mean age at diagnosis was 56.4 years for the ID group and 41.8 years for the NID group (P < .001). The ID group was 54.8% male compared with 13.5% in the NID group (P < .001). The ID group had higher stage disease compared with the NID group (P = .003). There was no difference in tumor size (P = .91), invasion (P = .76), lymph node involvement, or distant metastases (P > .99).

Conclusions and Relevance  Patients with ID thyroid cancer tend to be older at presentation, have higher stage disease, and are more likely to be male compared with patients with NID thyroid cancer. There does not appear to be a significant difference in the size, pathology, or behavior of the tumor at presentation between ID and NID thyroid cancers. These findings imply that improved detection may not represent the only cause of the increased incidence of thyroid cancer.

The incidence of thyroid cancer has nearly tripled in the past 30 years according to Surveillance, Epidemiology and End Results data. The incidence of thyroid cancer was 4.9 per 100 000 in 1975 vs 14.3 per 100 000 in 2009.¹ The mortality from thyroid cancer has remained quite stable, however, at approximately 0.5 per 100 000.¹ The reason behind the increasing incidence of thyroid cancer remains unclear.

Some authors have attributed this increase in incidence to the improved sensitivity of diagnostic techniques and imaging modalities, allowing for diagnosis of subclinical thyroid cancers.² Others have argued that these findings represent a true rise in incidence. Specifically, several studies have claimed that heightened diagnostic scrutiny does not fully explain the upward trend because there is also a statistically significant increase in large-sized thyroid cancers.^3-5 Instead, the elevation in incidence may be secondary to yet undetermined environmental or lifestyle factors.

Although several studies have evaluated patient demographics and size of cancer at time of presentation for all thyroid cancers during the past 30 years, to our knowledge, no studies have compared these same characteristics in incidentally discovered (ID) and nonincidentally discovered (NID) thyroid cancers.^2-6 To this effect, the aim of this study is to compare the clinical and pathologic characteristics of ID thyroid cancers via non–thyroid-related imaging modalities with NID thyroid cancers. If the rise in incidence is apparent secondary to increased detection of subclinical disease, it would be expected that these ID thyroid cancers would be smaller and present at an earlier stage in the disease process.^2-4 However, if the observed elevation in incidence of thyroid cancer is indeed reflective of the disease burden, then little difference in size and stage at presentation should be evident between the groups.

This study was approved by the institutional review board at the Pennsylvania State University–Milton S. Hershey Medical Center. Patients were identified from July 1, 2008, through June 30, 2009, by cross-referencing the International Classification of Diseases, Ninth Revision, Clinical Modification billing code (193) for thyroid cancer with the Current Procedural Terminology code (76536) for thyroid ultrasound. This allowed us to identify patients with a diagnosis of thyroid cancer who were under active surveillance during this period. In total, 263 patients with thyroid cancer were identified. Twenty-five were excluded due to incomplete medical records. Of the remaining patients, 31 and 207 were determined to have ID and NID thyroid cancer, respectively (Figure). For this study, ID thyroid cancer was defined as a nodule initially discovered on imaging not performed specifically for evaluation of the thyroid. The NID thyroid cancers were classified as such when the initial nodules were first discovered on a clinical examination, on imaging for evaluation of the thyroid or for symptoms that could be related to thyroid enlargement, such as dysphagia, or through the workup of any thyroid endocrinopathy. A retrospective medical record review was performed to extract age at time of diagnosis, sex, focality, pathology, size of dominant nodule at the time of diagnosis, presence of capsular or lymphovascular invasion or both, and presence of regional or distant metastasis. Staging using the TNM and AMES (age, metastases, extent, and size) systems was performed for each patient. Statistical analysis was performed using R, version 2.15.2 (R Development Core Team). Two-tail sample t tests and Fisher exact tests were used where appropriate, with a P  < .05 as the predetermined threshold for significance.

A total of 238 patients (31 patients with ID thyroid cancer and 207 patients with NID thyroid cancer) were examined. For the ID thyroid cancer group, the initial imaging modality and reason for obtaining the study are listed in Table 1. Approximately 55% of the ID thyroid cancers were discovered on computed tomography (CT) scan, 19.4% via ultrasound, 16.1% via positron emission tomography (PET) imaging, and 6.5% by magnetic resonance imaging (MRI) (Table 2).

A significant difference was noted in the mean age at diagnosis and sex between the 2 groups (Table 3). The mean age at diagnosis was 56.4 years for the ID group and 41.8 years for the NID group (P < .001). Of note, 77.4% of the ID group and 39.6% of the NID group was older than 45 years. Men constituted 54.8% of the ID group and 13.5% of the NID group (P < .001). The odds of being a male in the ID group were 7.66 times as high as the odds of being a male in the NID group (95% CI, 3.17-18.91).

A significant difference was noted in the stage of thyroid cancer at presentation when using the TNM staging system (Table 4). The ID group was noted to have an overall higher stage at presentation, with 48.4% having stage I disease, 16.1% having stage II disease, and 29.0% having stage III disease (P = .003). Most patients in the NID group (76.8%) had stage I disease. The AMES staging system did not reveal a significant difference between the 2 groups, but again the ID group had a higher percentage of patients (22.6% vs 12.6%) with high-risk disease (P = .15).

In terms of tumor pathology, there was no significant difference between the 2 groups (Table 5). Most cancers were papillary thyroid cancers, at a rate of 83.9% and 87.9% for the ID and NID groups, respectively (P = .40). The second most common diagnosis for both groups was follicular cancer followed by Hürthle cell and medullary thyroid cancer.

There was no significant difference in the size or focality of the cancer at the time of presentation between the 2 groups (Table 5). The mean size of the tumor was 2.15 cm for the ID group and 2.11 cm for the NID group (P = .91). The range in tumor sizes was 0.5 to 9 cm in the ID group and 0.1 to 8.7 cm in the NID group. In addition, there was no difference in size distribution between the 2 groups when using the TNM staging guidelines for tumor size (P = .26) (Table 6). Of the multifocal cancers, 32.3% were in the ID group and 44.4% were in the NID group (P = .24).

Differences in the presence of local invasion, as well as regional and distant metastases, between the 2 groups were also not significant (Table 5). Local invasion was absent in 67.7% of ID and 60.9% of NID thyroid cancer cases (P = .76). At the time of diagnosis, 22.6% in the ID group and 20.8% in the NID group had lymph node metastases (P  >.99). The ID group did not have any distant metastases, while in the NID group, 1 patient had distant metastases.

Numerous studies on ID thyroid nodules, or thyroid “incidentalomas,” have focused on their prevalence in specific imaging modalities and the malignant carcinoma rates of these lesions.^7-10 Reported prevalence rates of incidentalomas are approximately 16% for CT and MRI, 9.4% for carotid duplex, and 2% to 3% for PET and CT scans.^7,8 When grouped together, the reported malignant carcinoma rate of thyroid incidentalomas ranges between 15% and 24%, but this varies depending on the imaging modality used for discovery.^8,11 For incidental thyroid nodules found by CT and MRI, reported malignant carcinoma rates range from 3.9% to 11.3%.^7,12-14 Thyroid incidentalomas on PET imaging have reported malignant carcinoma rates ranging from 13.6% to 62.5%.^9,12,15,16 Some studies have looked at thyroid incidentalomas as diagnostic dilemmas, discussing approaches for patients with an incidental thyroid lesion.^10,12,17 Other studies have investigated thyroid cancers discovered incidentally via surgical pathology following thyroidectomy for presumed benign thyroid disease. These studies suggest that pathologically incidental thyroid cancers are biologically less aggressive compared with known cancers prior to surgical removal, with more favorable prognostic factors, such as size, capsular invasion, and lymph node involvement.^18,19 However, to our knowledge, no studies have analyzed the characteristics of radiologically incidental thyroid cancers in comparison with nonincidental thyroid cancers.

We found no significant difference between the tumor characteristics and tumor behavior of ID and NID thyroid cancers. Specifically, pathology, tumor size, rate of local invasion, and rate of regional and distant metastasis did not differ between ID and NID thyroid cancers. Notably, there was no significant difference in the size distribution of thyroid cancers between the groups. If heightened diagnostic scrutiny was causing the increased incidence of thyroid cancer, one would expect ID thyroid cancers to be smaller and less advanced in stage at the time of presentation. Our results support those of authors who have reviewed Surveillance, Epidemiology and End Results data and similarly shown an increase in all size distributions of thyroid cancer at the time of diagnosis.^3-5

Age, an important and well-known prognostic indicator in thyroid cancer, is incorporated in multiple staging systems for thyroid cancer, including AMES, MACIS (metastasis, age at presentation, completeness of surgical resection, invasion, size), and TNM.^20,21 A recent study investigating prognostic factors for life expectancy for differentiated thyroid cancer showed reduced life expectancy in patients aged 45 to 59 years and 60 years and older, while those aged 30 to 44 years showed a minimal decrease in life expectancy.²² We found a significant age difference at the time of diagnosis between ID and NID thyroid cancers. The mean age at diagnosis in the ID group was 56.4 years compared with 41.8 years in the NID group. One explanation for the older age in the ID group is that use of medical services rises with age, leading to increased investigational studies responsible for identifying incidental thyroid lesions. Since advanced age is inherently a risk factor for malignant carcinomas, elderly patients in whom an incidental thyroid lesion is discovered may be subject to more aggressive diagnostic pursuit than younger individuals, leading to possible selection of older patients in the ID group.^12,17

Interestingly, the ID group had an overall higher stage vs the NID group when compared using the TNM staging system. The ID group had less stage I disease than did the NID group and significantly more stage II and III disease. While there was no difference in tumor characteristics, such as size, invasion, lymph node, and distant metastases, a significant difference was noted in the age at presentation between the 2 groups. Specifically, a higher percentage of patients older than 45 years was in the ID group (77.4%) compared with the NID group (39.6%). Given that age older than 45 years is a staging parameter in the TNM system, this finding is likely responsible for the increased overall staging at presentation in the ID group. Using the AMES staging system, which incorporates both sex and age, we noted that a higher percentage of patients in the ID group had high-risk disease (22.6%) compared with the NID group (12.6%). These findings did not reach statistical significance, however, likely secondary to our small sample size.

There was also a notable difference in sex distribution between the 2 groups. Typically, the rate of thyroid cancer is 3-fold higher in females than in males.²³ It has been proposed that the reason for this wide disparity could be secondary to the influence of estrogen, but no clear correlation exists at this time.²³ We noted a higher percentage of males in the ID group compared with the NID group, at 54.8% vs 13.5%, respectively. One explanation for this discrepancy is that women are more likely to use medical services during their lifetime, especially during reproductive and perimenopausal years, leading to increased surveillance for and, hence, discovery of thyroid disease.²⁴ Men, in contrast, are less likely to seek routine medical care, which may lead to delayed diagnosis of a thyroid lesion until medical care is more necessary, such as following a traumatic event or at an advanced age. The odds of being male in the ID group were statistically higher compared with the NID group. Male sex is another known independent risk factor in thyroid cancer, with men having a worse prognosis than women with similarly staged disease.²⁵

Nonetheless, it is difficult to ignore the contribution of advanced imaging techniques and the increased use of diagnostic studies on the discovery of incidental thyroid nodules. Studies have shown that ID thyroid cancer accounts for 11% to 15% of thyroid malignant carcinomas.^26,27 Use of imaging modalities such as CT, MRI, and ultrasound is more prevalent, with a reported annual percentage increase of 6.7%, 11.2%, and 7.2%, respectively, between 1992 and 2001 in the Medicare population.²⁸ It is likely that a combination of a true rise in the incidence of thyroid cancer as well as an escalation in diagnosis is responsible for the observed increased incidence of thyroid cancer in recent decades.

Our study does have some limitations. Notably, our sample size is limited. There is also a large difference in the number of patients identified within each group, with 31 patients in the ID group and 207 patients in the NID group. Nonetheless, our conclusions are based on statistically significant results. Another limitation of this study is that we do not address all incidental nodules at our institution in this study design, but we plan to augment our database to include this information. In addition, medical comorbidities were not taken into account, and our data may be influenced by confounding factors that have not been identified.

The aim of this study is to compare the characteristics of ID and NID thyroid cancer. The ID thyroid cancers were diagnosed at an increased age compared with NID thyroid cancers. In addition, men were more likely to have an ID thyroid cancer than women. Both of these patient characteristics are associated with a worse prognosis in thyroid cancer. Consequently, the ID group had higher stage disease according to the TNM staging system. The implication is that ID thyroid cancer may have a worse prognosis, but further studies are warranted.

We also demonstrated that there were no significant differences in tumor size, invasion, and metastases to suggest that the overall rise in incidence in thyroid cancer is due to increased diagnosis of subclinical disease. This may imply that the overall elevation in the incidence of thyroid cancer is not solely due to the heightened diagnosis of incidental cancer and may represent a true rise in incidence. Other lifestyle and environmental factors that may be contributing to the increased incidence of thyroid cancer require further investigation.

Corresponding Author: David Goldenberg, MD, Division of Otolaryngology–Head and Neck Surgery, Department of Surgery, The Pennsylvania State University–Milton S. Hershey Medical Center, 500 University Dr, Mail Code H091, Hershey, PA 17033-0850 (dgoldenberg@hmc.psu.edu).

Submitted for Publication: February 9, 2013; final revision received May 8, 2013; accepted May 22, 2013.

Published Online: October 10, 2013. doi:10.1001/jamaoto.2013.5050.

Author Contributions: Mr Yoo and Drs Chaikhoutdinov and Goldenberg had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Yoo, Chaikhoutdinov, Mitzner, Goldenberg.

Acquisition of data: Yoo, Chaikhoutdinov, Mitzner.

Analysis and interpretation of data: Yoo, Chaikhoutdinov, Liao, Goldenberg.

Drafting of the manuscript: Yoo, Chaikhoutdinov, Liao, Goldenberg.

Critical revision of the manuscript for important intellectual content: Yoo, Chaikhoutdinov, Mitzner, Goldenberg.

Statistical analysis: Yoo, Chaikhoutdinov, Liao.

Administrative, technical, and material support: Mitzner, Goldenberg.

Study supervision: Chaikhoutdinov, Mitzner, Goldenberg.

Conflict of Interest Disclosures: None reported.