New guidance, technology on the horizon for thyroid nodule risk assessment

We were unable to process your request. Please try again later. If you continue to have this issue please contact

Key takeaways:

  • Upcoming changes to thyroid nodule guidance could help further pinpoint cancer risk for patients.
  • AI and radiomics may someday assist providers in better determining malignancy.

NEW ORLEANS —New risk stratification guidance for classifying thyroid nodules may come soon as imaging technology continues to evolve, according to a speaker at the AACE annual meeting.

The quality of thyroid ultrasound imaging has improved dramatically since its first use more than 50 years ago, according to Jennifer A. Sipos, MD, professor of medicine in the division of endocrinology and metabolism at the Ohio State University. As imaging has improved, recommendations on the management of thyroid nodules have changed over the years. Sipos looked ahead at what’s next in the field with a preview of new draft guidance from the American Thyroid Association on thyroid nodule risk stratification and how new technology could change transform risk assessment.

Thyroid ultrasound female 2 2019.

Advances in thyroid ultrasound have led to changes in thyroid nodule risk assessment recommendations over the years. Image: Adobe Stock

Impact of risk stratification

Improvements in ultrasound technology early this century led to a rise in the number of thyroid ultrasounds performed, which led to an increase in thyroid cancer diagnoses. However, Sipos noted, most of the new diagnoses were for nodules measuring 2 cm or less, and thyroid cancer mortality rates did not rise.

“As we introduced more ultrasounds, we diagnosed more cancers, but we really didn’t impact patient outcomes,” Sipos said during a presentation.

Jennifer A. Sipos

The increase in diagnoses plateaued after several organization published guidance to refine risk assessment for thyroid nodules. In 2016, the ATA published guidelines, and AACE partnered with the Associazione Medici Endocrinologi to release its own guidance,. In 2017, the American College of Radiology released the Thyroid Imaging Reporting and Data System (TI-RADS), which introduced a point system for categorizing all types of thyroid nodules.

A study published in Surgery in 2023 revealed the impact of the new guidelines. Researchers found the percentage of thyroid nodules categorized as Bethesda II decreased from 49.6% in 2014 to 19.4% in 2021, whereas the prevalence of Bethesda III nodules increased from 21.3% in 2014 to 51.5% in 2021.

“This is evidence that we are paying attention to these risk stratification systems and we’re not biopsying very low-risk nodules anymore,” Sipos said. “We’re content with following them, because we know the risk of malignancy is low.”

Changes to ATA risk stratification

The ATA is planning to unveil new thyroid nodules guidance in the near future, Sipos said. She presented a preview of the draft guidance, which included several changes.

One change is the guideline will classify all thyroid nodules, including those with calcification and partially cystic nodules. The guidance may also alter the risk for malignancy classifications to eliminate gaps. The draft presented by Sipos has a low risk for malignancy defined a 3% to 20% likelihood, an intermediate risk labeled as 20% to 50% likelihood and high risk as more than a 50% likelihood of malignancy. The new guidance will also be updated with current literature.

“There’s been no change in the very low and benign [suspicion] groups,” Sipos said. “The biggest change has been in the intermediate suspicion group. In the high suspicion group, we added cystic papillary thyroid cancers, and we moved the non-markedly hypoechoic nodule with no suspicious features from intermediate risk down to low risk.”

The draft guidance includes a sonographic pattern system algorithm where the composition of nodules is defined as either solid, cystic or consistency cannot be assessed. For solid nodules, the number of suspicious features plays a major role in determining risk for malignancy.

For partially cystic nodules, those featuring a high-risk cystic appearance indicative of papillary cystic thyroid cancer are considered high suspicion, with all others considered low suspicion. The draft guidance adds a section for determining risk for spongiform, or microcystic, nodules. Spongiform nodules with macrocalcification or an irregular outer nodule margin are considered low suspicion, with other spongiform nodules classified as very low suspicion.

For nodules whose consistency cannot be assessed, the draft guidance states these should be classified as high suspicion if there is peripheral calcifications with protruding soft tissue, intermediate suspicion with irregular peripheral calcifications and no protruding soft tissue, and low suspicion with thin, regular peripheral calcifications and no protruding soft tissue.

Because the guidance is still being edited, Sipos said, changes could be made prior to publication, which may occur later in 2024.

Looking ahead

Risk assessment of thyroid nodules is poised to advance even more with AI and new technologies. Sipos said she believes AI will be used in evaluating ultrasound imaging in the future and will be used in collaboration with sonographers.

“I think it’s really an important tool to train other sonographers or maybe even a tool to use where there aren’t expert sonographers,” Sipos said. “But I do think there could be some very good technology for this in our world.”

Radiomics — the use of advanced quantitative imaging techniques for features of a nodule that the human eye cannot see — may represent another technologic advance in thyroid nodule risk assessment, according to Sipos.

Advances are also making their way into fine-needle aspiration. Sipos discussed a mixed-reality simulator being used at the University of Florida where trainees can learn how to perform fine-needle aspiration using a ballistic rejuvenable model that is molded into a human shape. Virtual reality shows the trainee how they are performing and gives feedback as they inject the needle into the model.


Ramonell KM, et al. Surgery. 2023;doi:10.1016/j.surg.2022.06.061.

Source link

error: Content is protected !!