The vast majority of papillary thyroid cancers (PTCs) diagnosed in the US between 1991 and 2019 were cases of overdiagnosis, with the detection and treatment of the cancers having no impact on population-based mortality, new research showed.

“Even accounting for a true underlying increase in the risk of developing thyroid cancer over [the study’s] time frame, our estimates confirmed that the detection of clinically inactive cancers is the primary factor in the rapid increase in thyroid cancer over the past three decades,” reported the authors of the study, published in JAMA Network Open. 

While the incidence of thyroid cancer in the US has increased by a striking 250% since 1990, the majority of cases have been small, early-stage thyroid cancers, with mortality related to the cancers remaining stable, the authors explained.

Autopsy studies show that as many as 10% of men and women have PTCs detected after death that they were not aware of during life, therefore increases in referrals for ultrasound and imaging are considered to largely explain the rise in incidence.

Ultrasonography is easily accessible, and referrals are common, based on symptoms or physical examination findings or after incidental nodules are found through imaging for other purposes. Studies show that thyroid ultrasonography has increased fivefold in the US since 2000, while cross-sectional imaging has increased 10-fold.

Validated Simulation Model

To better quantify the magnitude of thyroid cancer overdiagnosis, the authors used a validated simulation model, the Papillary Thyroid Carcinoma Microsimulation Model, utilizing data from the National Cancer Institute Surveillance, Epidemiology, and End Results program on sex- and age-specific PTC incidence, stage distribution, and mortality between 1975 and 2019.

In the model, deaths occurring among patients diagnosed with PTC that occurred prior to a PTC-related death were considered to be overdiagnosed cases.

The results showed that between 1991 and 2019 as many as 72%-94% of PTC cases met the definition of overdiagnosis, and those cases did not have an effect on mortality in the population.

Overdiagnoses rates were three to four times higher among women than men (lower to upper bound range: 75%-95% vs 63%-90%), with those results consistent across most age groups.

The percentage of individuals who were overdiagnosed with PTC based on age and sex was relatively stable over the study’s nearly three decades.

The model estimated that reducing the use of referral for thyroid ultrasonography for nonpalpable nodules by 33% would have reduced the incidence of PTC detection by 17% in 2019. A 67% reduction in ultrasonography referral could reduce the incidence of PTC by 41%. Notably, these changes would result in a less than 0.1% change in overall mortality.

Furthermore, the model estimated that a complete elimination of ultrasonography use for nonpalpable indications would have the effect of reducing thyroid cancer incidence to levels observed in 1991.

“In all of these simulations, there was negligible implication for population mortality,” the authors wrote.

The study’s estimates of overdiagnoses are higher than those previously reported, due to the model’s methodology, they explained. “We used a microsimulation model, while other investigators compared the ratio of observed-to-expected thyroid incidence.”

Currently, there are no official guidelines regarding when to refer patients to ultrasound but considering the “cascade of decision-making” about biopsy and management that can be set off by ultrasonography, better guidance is needed regarding when to order the testing, “particularly for nonpalpable, asymptomatic thyroid nodules,” the study authors asserted.

Guideline Progress

Some positive steps in addressing the overdiagnosis problem have included guidance regarding when to biopsy, first author David O. Francis, MD, a professor in the Department of Otolaryngology-Head & Neck Surgery at University of Wisconsin – Madison, told Medscape Medical News. 

“Guidelines from both the American Thyroid Association (2015) and the American College of Radiology (2017) recommend that thyroid nodules 1 cm or less generally not be biopsied,” he noted.

“These steps have helped somewhat flatten the rate of diagnosis of small cancers, but there is still a lot of room for improvement, as our study shows.”

Lessons can also be learned from countries where overdiagnosis rates are notably lower — such as Netherlands, said Francis.

“Since 2007, [Netherlands] have only done fine needle aspiration for people with palpable thyroid nodules therefore incidentally detected nodules on CT and MRI are rarely evaluated,” he explained.

Other measures that could make a difference in improving rates in the US include parsing referral guidelines for the use of ultrasound, said Francis. In addition, conducting a clinical trial comparing currently used strategies for workup with strategies to instead monitor nodules over time “would provide evidence that could help people feel more comfortable leaving those nodules in place that show themselves to not be clinically active.”

Complexities of Real-World Practice

In an accompanying editorial, Erik K. Alexander, MD, Division of Endocrinology, Metabolism & Diabetes, Mass General Brigham, Harvard Medical School, Boston, noted that population-based assessments can get complicated when translated to the complexities of real-world practice.

For instance, data has shown no difference in oncologic risk between palpable and nonpalpable nodules, he said. Furthermore, “the performance and quality of an excellent physical examination has been devalued over recent decades, making a palpable thyroid nodule less common more so due to the skillset of the healthcare practitioner rather than the biology of the tumor.”

In addition, ultrasonography, which remains highly effective in assessing cancer risk with a thyroid nodule, “often leads to a critical next step so often overlooked and devalued — which is the process of clinician and patient shared discussion and medical decision-making,” he added.

Alexander agreed that many, if not most, thyroid nodules require limited, or no diagnostic assessment at all.

“But some do, and some thyroid cancers are dangerous.”

He called for “use of the toolbox of medical options that are available — but use them wisely,” he cautioned.

“Talk to the patient and realize the myriad of factors that can impact shared clinical decisions,” he urged.

“And select next steps judiciously, noting that the longitudinal understanding of the patient and their thyroid nodule may well prove more valuable in comparison with immediate next interventions.”

Commenting on the study, Anupam Kotwal, MD, an associate professor with the Division of Diabetes, Endocrinology & Metabolism and medical director of the Thyroid Tumor Program at the University of Nebraska Medical Center, in Omaha, Nebraska, noted the need to still consider the smaller number of truly concerning cancers.

“Overall, the significance of these findings is that they do demonstrate the overdiagnosis of thyroid cancer, especially that of indolent nature which would not have impacted lifespan,” he told Medscape Medical News.

“However, other studies have demonstrated increases in the risk of aggressive thyroid cancer, although not to the same extent, hence all the increase in thyroid cancer rates cannot be attributed to overdiagnosis alone.”

Kotwal noted further caveats in the suggestion of not referring nonpalpable nodules include the key concern that physicians can have varying expertise in neck examinations to identify nodules.

In addition, “the suggestion also does not take into account the change in the thyroid position to being more low lying in older age or difficulty in palpating moderate size thyroid nodules in those with more muscular or thicker neck, such as in men,” he added. “Hence not evaluating sonographically a nodule that has been identified by other imaging or in the setting of compressive neck symptoms may miss clinically significant thyroid cancer.”

The use of neck ultrasound in such cases “can serve as a low cost, low risk tool in the clinician armamentarium to then make a decision after patient-centered discussion about next steps for a thyroid nodule,” Kotwal said.

Ultimately, “clinicians should be aware of the overdiagnosis of thyroid cancer and hence pursue ultrasonography for appropriate indications but should not use palpable nature alone as a gatekeeper before further evaluation,” he added.

“They should utilize patient-centered discussion to guide the next steps, including, but not limited to, neck ultrasonography, biopsy, and then a discussion regarding surveillance vs surgery.”

This study received funding from the National Cancer Institute/National Institutes of Health. Alexander reported receiving consulting fees from Roche Diagnostics Inc, Regeneron Inc, and Veracyte Inc. Kowtal reports serving as a consultant for Veracyte, Inc..