Fax: (404) 329-6450
Increasing incidence of differentiated thyroid cancer in the United States, 1988–2005
Article first published online: 13 JUL 2009
Copyright © 2009 American Cancer Society
Volume 115, Issue 16, pages 3801–3807, 15 August 2009
How to Cite
Chen, A. Y., Jemal, A. and Ward, E. M. (2009), Increasing incidence of differentiated thyroid cancer in the United States, 1988–2005. Cancer, 115: 3801–3807. doi: 10.1002/cncr.24416
- Issue published online: 3 AUG 2009
- Article first published online: 13 JUL 2009
- Manuscript Accepted: 29 DEC 2008
- Manuscript Revised: 12 DEC 2008
- Manuscript Received: 3 SEP 2008
- annual percentage change;
- fine-needle aspiration;
- follicular cancer;
- papillary cancer;
- thyroid cancer;
Studies have reported an increasing incidence of thyroid cancer since 1980. One possible explanation for this trend is increased detection through more widespread and aggressive use of ultrasound and image-guided biopsy. Increases resulting from increased detection are most likely to involve small primary tumors rather than larger tumors, which often present as palpable thyroid masses. The objective of the current study was to investigate the trends in increasing incidence of differentiated (papillary and follicular) thyroid cancer by size, age, race, and sex.
Cases of differentiated thyroid cancer (1988-2005) were analyzed using the National Cancer Institute's Surveillance Epidemiology and End Results (SEER) dataset. Trends in incidence rates of papillary and follicular cancer, race, age, sex, primary tumor size (<1.0 cm, 1.0-2.9 cm, 3.0-3.9 cm, and >4 cm), and SEER stage (localized, regional, distant) were analyzed using joinpoint regression and reported as the annual percentage change (APC).
Incidence rates increased for all sizes of tumors. Among men and women of all ages, the highest rate of increase was for primary tumors <1.0 cm among men (1997-2005: APC, 9.9) and women (1988-2005: APC, 8.6). Trends were similar between whites and blacks. Significant increases also were observed for tumors ≥4 cm among men (1988-2005: APC, 3.7) and women (1988-2005: APC, 5.70) and for distant SEER stage disease among men (APC, 3.7) and women (APC, 2.3).
The incidence rates of differentiated thyroid cancers of all sizes increased between 1988 and 2005 in both men and women. The increased incidence across all tumor sizes suggested that increased diagnostic scrutiny is not the sole explanation. Other explanations, including environmental influences and molecular pathways, should be investigated. Cancer 2009. © 2009 American Cancer Society.
The incidence of thyroid cancer has been increasing in the United States since 1980. Such increases in incidence rates can reflect either a true increase in the occurrence of the specific cancer or increased detection of the cancer. One example of the latter is prostate cancer, in which a marked increase in incidence was noted after the widespread use of prostate-specific antigen (PSA). Some researchers have attributed the increase in the incidence of thyroid cancer solely or in large part to an increase in the diagnosis of small tumors1 and have suggested that it is safe to observe most patients who have thyroid nodules that measure <1 cm.2 Evidence-based guidelines from the American Thyroid Association recommend no needle biopsies unless the patient has multiple nodules and/or unless the lesion measures >1 cm. The objective of the current study was to investigate thyroid cancer incidence rates by sex, race, size, and Surveillance, Epidemiology, and End Results (SEER) stage to determine whether the increasing incidence of thyroid cancer is confined solely to the diagnosis of smaller tumors.
MATERIALS AND METHODS
Cases of differentiated thyroid cancer diagnosed between 1988 and 2005 were selected from the National Cancer Institute's SEER cancer registry using International Classification of Diseases for Oncology codes for papillary and follicular cancers (codes 8050, 8260, 8290, 8330-8331, 8335, and 8340-8344, as outlined in the sixth edition of the American Joint Committee on Cancer Cancer Staging Manual3). Joinpoint analysis was performed by age (all ages, aged <45 years, and aged ≥45 years), race (all, white, and black), sex (men, women), tumor size (all sizes, <1 cm, 1-1.9 cm, 2-2.9 cm, 3-3.9 cm, and ≥4 cm), and SEER stage (localized, regional, distant). The annual percentage change (APC) in incidence was calculated from the joinpoint model.
In total, 30,766 cases of differentiated thyroid cancer were diagnosed in all SEER registries between 1988 and 2005 (Table 1). The vast majority of cases occurred among whites (n = 25,461; 83%) and among individuals aged <45 years (n = 14,746; 48%), 23,308 cases (76%) occurred among women, and 14,746 cases (48%) occurred among individuals aged <45 years. With regard to tumor size, 7571 tumors (25%) measured <1.0 cm, 12,896 (42%) measured from 1 cm to 2.9 cm, 2893 (9%) measured from 3 cm to 3.9 cm, 3412 (11%) measured ≥4 cm, and 3818 (12%) had unknown measurements. Localized SEER stage (Table 2) formed the majority of SEER-staged cases (n = 18,373; 60%).
|Sex/Race||No.||Joinpoint Analyses, 1988-2005†|
|Trend 1||Trend 2|
|Years||APC (95% CI)‡||Years||APC (95% CI)‡|
|All tumor sizes|
|All races||30,766||1988-1998||3.5 (2.7-4.2)§||1998-2005||6.7 (5.8-7.7)§|
|White||25,461||1988-1999||4 (3.3-4.6)§||1999-2005||7.7 (6.4-8.9)§|
|Aged ≤45 y|
|All races||14,746||1988-2005||4.4 (3.9-4.9)§|
|Aged ≤45 y|
|All races||16,020||1988-1998||3.4 (2.5-4.4)§||1998-2005||7.6 (6.3-8.9)§|
|White||13,198||1988-1999||4 (3.1-5.0)§||1999-2005||8.9 (7.3-10.6)§|
|All races||7458||1988-1999||2.9 (1.9-3.9)§||1999-2005||6.3 (4.4-8.2)§|
|White||6402||1988-1999||3.2 (2.2-4.3)§||1999-2005||6.9 (4.8-9.0)§|
|Black||347||1988-2005||2.6 (−0.5 to 5.9)|
|Aged <45 y|
|All races||2780||1988-2005||3.5 (2.4-4.6)§|
|Black||116||1988-2005||2.8 (−1.9 to 7.8)|
|Aged ≥45 y|
|All races||4678||1988-1999||3.2 (2.2-4.2)§||1999-2005||6.5 (4.6-8.4)§|
|White||4007||1988-1999||3.6 (2.7-4.5)§||1999-2005||7.1 (5.5-8.8)§|
|Black||231||1988-2005||2.6 (−0.6 to 5.9)|
|All races||23,308||1988-1998||3.7 (2.7-4.8)§||1998-2005||7 (5.6-8.5)§|
|White||19,059||1988-1999||4.3 (3.3-5.3)§||1999-2005||8 (6.2-9.9)§|
|Aged <45 y|
|All races||11,966||1988-2005||4.7 (4.2-5.2)§|
|Aged ≥45 y|
|All races||11,342||1988-1998||3.6 (2.1-5.1)§||1998-2005||8.3 (6.4-10.2)§|
|White||9191||1988-1999||4.2 (2.8-5.7)§||1999-2005||9.8 (7.2-12.5)§|
|Sex/Size/Stage||No.||Joinpoint Analyses: 1988-2005†|
|Trend 1||Trend 2|
|Years||APC (95% CI)‡||Years||APC (95% CI)‡|
|Tumor size, cm|
|<1.0||1520||1988-1997||4 (0.8-7.3)§||1997-2005||9.9 (7.3-12.7)§|
|Unknown||1167||1988-1997||3.3 (1.0-5.5)§||1997-2005||−8 (−10.4 to −5.5)§|
|Localized||3792||1988-1999||4.1 (2.7-5.5)§||1999-2005||9.1 (6.6-11.7§)|
|Unstaged||182||1988-2005||−2.7 (−5.7 to 0.4)|
|Tumor size, cm|
|1.0-2.9||10,289||1988-1994||0.4 (−3 to 3.8)||1994-2005||6.3 (5.3-7.4)§|
|Unknown||2651||1988-2000||1.1 (−0.9 to 3.1)||2000-2005||−9.9 (−16.7 to −2.5)§|
|Regional||7443||1988-1992||−2.7 (−7.2 to 2.1)||1992-2005||4.3 (3.6-4.9)§|
|Unstaged||530||1988-2005||−3.6 (−6.6 to −0.5)§|
The age-adjusted incidence of thyroid cancer was 3 times higher among women than among men (Fig. 1); however, increases in incidence were observed in both sexes. The mortality rates remain unchanged during the study period (Fig. 1) despite higher rates of diagnoses. Among women, the age-adjusted incidence (cases per 100,000) of differentiated thyroid cancer increased from 6.4 per 100,000 in 1988 to 14.9 per 100,000 in 2005. Among all women, an increasing trend in incidence (APC, 3.7; P < .001) was observed from 1988 through 1998, and a much more rapidly increasing trend (APC, 7; P < .001) was observed from 1998 through 2005. The rates increased for both white women and black women during this study period (Table 1). Among all women aged <45 years, the incidence of thyroid cancer rose from 5.0 per 100,000 in 1988 (Fig. 2) to 10.7 per 100,000 in 2005 (APC, 4.7; P < .001) (Table 1). Among all women aged ≥45 years, the incidence of thyroid cancer rose from 8.8 per 100,000 in 1988 to 22.9 per 100,000 in 2005 (1988-1998: APC, 3.6 [P < .001]; 1998-2005: APC, 8.3 [P < .001]). The trends for white women and black women were the same for those aged <45 years and those aged ≥45 years (Table 1). Although the largest increase in incidence among women of all ages was observed for tumors that measured <1 cm (1988-2005: APC, 8.6 [P < .001]) (Table 2), significant increases also were observed for tumors that measured from 1.0 cm to 2.9 cm (1994-2005: APC, 6.3 [P < .001]), from 3.0 cm to 3.9 cm (1988-2005: APC, 4.4 [P < .001]), and ≥4 cm (1988-2005: APC, 5.7 [P < .001]). Among all women, according to SEER staging, the incidence rates also increased for localized stage (1988-2005: APC, 6.7 [P < .001]), regional stage (1992-2005: APC, 4.3 [P < .001]), and distant stage (1988-2005: APC, 2.3 [P <.001]).
Among men, the age-adjusted incidence of well differentiated thyroid cancer increased from 2.5 per 100,000 in 1988 to 5.1 per 100,000 in 2005 (Fig. 1). Again, mortality rates among men remain unchanged (Fig. 1). Similar to the pattern among women, there was an increasing incidence trend among all men (1988-1999: APC, 2.9 [P < .001]; 1999-2005: APC, 6.3 [P < .001]). The rate of change was lower among black men (1988-2005: APC, 2.6 [P < .001]) than among white men (Table 1). Among all men aged <45 years, the incidence of thyroid cancer rose from 1.4 per 100,000 in 1988 (Fig. 2) to 2.5 per 100,000 in 2005 (1988-2005: APC, 3.5 [P < .001]) (Table 1). Among all men aged ≥45 years, the incidence of thyroid cancer rose from 4.6 per 100,000 in 1988 to 9.9 per 100,000 in 2005 (1988-1999: APC, 3.2[P < .001]; 1999-2005: APC, 6.5 [P < .001]). The increasing trends for white and black men aged <45 years were significant, whereas the trends for black men aged ≥45 years were not significantly increased (Table 1). Although the largest increase in incidence among all men was observed for tumors that measured <1 cm (1997-2005: APC, 9.9 [P < .001]) (Table 2), significant increases also were observed for tumors that measured from 1.0 cm to 2.9 cm (1988-2005: APC, 5.5 [P < .001]), from 3.0 cm to 3.9 cm (1988-2005: APC, 2.7 [P < .001]) and ≥4 cm (1988-2005: APC, 3.9 [P < .001]). Among all men, according to SEER staging, the incidence rates also increased for localized stage (1988-1999: APC, 4.1 [P < .001]; 1999-2005: APC, 9.1 [P < .001]), regional stage (1988-2005: APC, 2.4 [P < .001]), and distant stage (1988-2005: APC, 3.7 [P < .001]).
Figure 3 compares incidence curves for women and men segregated by tumor size. Women had twice the number of cases; however, the trends were similar across sex when analyses were stratified by tumor size.
Increasing incidence rates can reflect either a true increase in the number of cancers or increased detection. We undertook the current study to explore patterns of differentiated thyroid cancer incidence by size, age, race, and sex to determine whether the increased incidence occurred primarily in smaller tumors, which would be expected if the increase were related primarily to increased detection. Our principal finding was that the greatest increase was for occult tumors; however, an increased in thyroid cancer of all sizes also was observed, including those >4 cm and those with distant disease (SEER stage).
Our observations are in contrast to other reports that evaluated trends in thyroid cancer by tumor characteristics, including tumor size. One such study presented descriptive data on incidence trends by tumor size but did not use statistical modeling techniques and concluded that the increase in incidence rate among thyroid cancers was attributable to the detection of smaller tumors.1 Those authors believed that this increase resulted from more frequent use of ultrasonography in locating thyroid nodules and the subsequent use of fine-needle aspiration to ascertain pathology. This practice of finding and biopsying small nodules is controversial, because the natural history of small thyroid cancers may not be important clinically; thus, there may be a bias to over treat (ie, operate) small, clinically insignificant thyroid nodules. A study of thyroid cancer incidence in Lithuania using joinpoint analysis also concluded that the increase in thyroid cancer incidence was attributable largely to increases in the application of ultrasound-guided fine-needle aspiration biopsy in clinical practice.5 Another group reported that more patients who resided in zip codes with a higher median income and with private insurance were diagnosed with thyroid cancer.6 This finding again supports the hypothesis that more fine-needle diagnoses are occurring, thus contributing to the increasing incidence of early thyroid cancer.
In contrast to a recent article by Morris et al, in our study, we observed similar increasing trends for white and black individuals.6 Morris et al reported a divergence of incidence rates between whites and blacks after the late 1990s.6 Those authors also examined the correlation between insurance status and incidence rates and reported a decrease in thyroid cancer diagnosis among uninsured individuals, suggesting again that the increase in diagnosis was predominantly because of the diagnosis of smaller tumors by improved screening with ultrasound, fine-needle biopsy, and other radiologic interventions mostly available to those with insurance. In our study, only black men aged ≥45 years did not demonstrate a significant APC in incidence rates (they did demonstrate an increase, but it was not statistically significant). Black women aged ≥45 years did demonstrate a significant APC, and so did all blacks combined (men and women collectively).
In addition, Correa and Chen7 also reported a population-based analysis of thyroid cancer. In both the report by Correa and Chen and the report by Morris et al, the methodology differed from ours, in that the APC was not calculated. Correa and Chen also demonstrated an increasing incidence trend for their study period, although it was more gradual than that demonstrated in the current study period from 1988 to 2007. Their article also reported lower incidence for black men, similar to our findings of an insignificant APC for black men aged ≥45 years.
Our finding that even the largest tumors and tumors with distant spread of disease had an increasing incidence trend suggests that other reasons for this increase, including environmental, dietary, and genetic causes, need to be explored. To our knowledge, there is no new evidence to suggest that the exposure of human beings to radiation, a well known environmental risk factor, has increased over time to account for the observed increase in thyroid cancer.
Dietary influences also have been explored. In 2002, Mack et al reported a decreased risk of thyroid cancer with the consumption of fresh fruits and vegetables.8 In 2003, Markaki et al9 also reported an association of lower risk for thyroid cancers with dietary patterns of fruits, raw vegetables, and mixed raw vegetables and fruits. In contrast, a dietary pattern of fish and cooked vegetables led to an increased risk of follicular cancer. Dijkstra et al reported a correlation between increasing thyroid cancer incidence and increasing dietary iodine intake.10 None of those reports indicated whether these dietary patterns of consumption had changed significantly from 1988 to 2005.
Exploration of molecular pathways, such as the phosphoinositide 3-kinase/protein kinase B (PI3K/Akt pathway), also has been performed. Larson et al reported that patients with Hashimoto thyroiditis (a condition in which PI3K/Akt expression increased) were 3 times more likely to develop thyroid cancer (in which PI3K/Akt expression also is increased). This molecular link isolated to Hashimoto thyroiditis suggests that there is a molecular mechanism for thyroid carcinogenesis.11 Other etiologies may include mutations in the v-raf murine sarcoma viral oncogene homolog B1 BRAF and rearrangements of the papillary thyroid carcinoma oncogene RET/PTC.12
In conclusion, the incidence of thyroid cancer has been increasing. In partial contrast to previously published reports, in our study, we observed that the increasing incidence of thyroid cancer is not only the product of greater detection of smaller tumors. Instead, the detection of all sizes of well differentiated thyroid cancers increased from 1988 to 2005. Despite better detection of these cancers, survival rates have not improved, suggesting the need for further exploration into the etiologies of increasing thyroid cancer rates.
Conflict of Interest Disclosures
The authors made no disclosures.
- 3GreeneFL, PageDL, FlemingID, et al, eds. AJCC Cancer Staging Manual. 6th ed. New York, NY: Springer-Verlag; 2002.
- 4Population Projections of the United States by Age, Sex, Race, and Hispanic Origin: 1995 to 2050, US Bureau of the Census, Current Population Reports, P25-1130. Washington, DC: US Government Printing Office; 1996..