A Comparison of Sonographic Thyroid Nodule Risk Stratification Systems in an Unselected Cohort

• Clin Thyroidol 2021;33:484–486.
• The evaluation of patients with thyroid nodules requires adequate stratification to estimate the risk of thyroid cancer:
  • Evaluation also requires an understanding of the patient’s situation
• To assist clinicians in this task, multiple thyroid nodule sonographic risk-stratification systems (RSSs) have been developed, including:
  • The American Thyroid Association (ATA) system
  • The American College of Radiology Thyroid Imaging Reporting and Data System (ACR TI-RADS)
  • The Korean Thyroid Imaging Reporting and Data System (K-TIRADS)
  • The European Thyroid Imaging Reporting and Data System (EU-TIRADS)
  • The Artificial Intelligence Thyroid Imaging Reporting and Data System (AI TI-RADS)
• In general, studies assessing the performance of these RSSs have been retrospective and have included patients who have undergone a diagnostic intervention (i.e., thyroid nodule biopsy or surgery):
  • This type of study design can introduce bias into the assessment of diagnostic properties by increasing the proportion of high risk cases
• Although RSSs for thyroid nodules aim to standardize thyroid cancer risk assessment and recommendations:
  • The ACR TI-RADS is associated with a lower number of recommendations for thyroid biopsy:
    • This is due, in part, to the system’s relatively more conservative thresholds for recommending thyroid biopsy
• In the current study, a comparison of thyroid nodule RSSs is evaluated:
  • In thyroid nodule risk distribution in an unselected group of patients undergoing thyroid ultrasound
  • By assessing the impact of risk distribution, in addition to the size thresholds, on biopsy recommendations
• Methods:
  • This was a multi-institutional study of seven radiology practices that participate in the ACR registry
  • Each practice prospectively submitted thyroid ultrasound reports on adult patients between October 2018 and March 2020
  • Sites provided the maximum size of the thyroid nodules and followed structured reporting according to the five ACR TI-RADS ultrasound features
  • Patients with thyroid cancer, nodule size less than 5 mm or greater than 5 cm, and incomplete / unrealistic data were excluded
  • The submitted reports were then retrospectively categorized following the criteria of ACR TI-RADS, the ATA system, K-TIRADS, EU-TIRADS, and AI TI-RADS
  • The distribution of risk categories and thyroid biopsy recommendation rates were compared
  • Because of the large sample size, even small differences in the proportions observed were expected to be statistically significant
• Results:
  • The study population consisted of:
    • 12,208 patients, mostly women (84%), encompassing a total of 27,933 thyroid nodules
  • The mean patient age was 60.7 years, and the mean (±SD) nodule size was 1.5 ± 0.92 cm
  • There were 1896 nodules that could not be classified according to the ATA system
  • According to ACR TI-RADS:
    • The distribution of thyroid nodules was:
      • TIRADS 1 (TR1) (3.1%), TR2 (8.2%), TR3 (31.0%), TR4 (48.3%), and TR5 (9.4%)
  • The ACR TI-RADS and AI TI-RADS:
    • Placed more nodules in the TR2 category (8.2% and 10%, respectively) than the other systems (range, 1.2%–2.5%)
  • EU-TIRADS:
    • Placed more nodules in the high suspicion / TR5-equivalent category (18.9%) than did the other systems (range, 9.1–12.5%)
  • AI TI-RADS had the lowest level of TR3 nodules (26.1%) as compared with the other systems (range, 31–44.4%)
  • In all, the most common category for all nodules was TR4 and its equivalents (42.1–48.3%)
  • ACR TI-RADS recommended biopsies in 29.1% of the thyroid nodules:
    • One of the lowest rates as compared with other systems (ATA, 58.7%; EU-TIRADS, 38.9%; K-TIRADS, 57%)
    • AI TI-RADS recommended biopsy for 26.3% of the nodules
  • Finally, when evaluating thyroid biopsy recommendations according to risk category:
    • The rate of recommended biopsy was similar in the TR5 categories and equivalents (68.7–75.5%)
  • There was variability for TR2:
    • With ACR TI-RADS recommending biopsy in 0% of the nodules (range for others, 2.8–17.7%)
  • The largest differences were for TR3 and TR4 nodules:
    • For which ACR TI-RADS and AI TI-RADS recommended biopsy in 19.0% to 22.3% and 32.7% to 33.7 %, respectively; the range for the other systems was 33.3 to 53.7% for TR3 and 29.0 to 64.0% for TR4
  • Conclusions:
    • Differences in the distribution of sonographic thyroid nodule risk categories and biopsy size thresholds among the various RSSs contribute to variability in clinical recommendations for thyroid biopsies
    • ACR TI-RADS generally recommends a lower number of biopsies:
      • Because of a combination of its risk assignment criteria and more conservative biopsy thresholds
  • This large multi-center study validates the clinical utility of thyroid nodule risk stratification:
    • By evaluating patients who underwent thyroid ultrasound regardless of their diagnostic workup
  • Using ACR TI-RADS, 11.3% of the thyroid nodules were considered either benign or very low risk for thyroid cancer, while 9.4% were considered high risk
  • Most patients had nodules considered either T3 or T4 (31.0% and 48.3%, respectively)
  • This distribution highlights the importance of high-quality ultrasound risk stratification in the evaluation of patients with thyroid nodules, as it can facilitate reassurance in patients with low-risk nodules
  • Similarly, a robust RSS may appropriately justify the consideration of biopsy in those at higher ris
  • In addition, the study evaluated the impact of size thresholds across the different risk categories to guide the need for thyroid biopsies
  • Application of ACR TI-RADS was associated with a biopsy rate of 29.1%:
    • Lower than that for all the other systems except for AI TI-RADS
  • This was driven, in part:
    • By fewer biopsy recommendations for nodules in categories TR3 and TR4
    • As well as a greater proportion of nodules in the TR2 category:
      • For which biopsy is not recommended
  • Taken together, these findings suggest that, in addition to different size thresholds for biopsy, the distribution of risk categories affects biopsy recommendation rates
  • In fact, the overall proportion of nodules recommended for biopsy varied between the systems, from 26.3% in AI TI-RADS to 58.7% in the ATA system
  • These findings highlight the urgent need for and value of current efforts to harmonize thyroid nodule RSSs, particularly in terms of risk-category definitions and management recommendations by malignancy estimates and size thresholds for biopsy
  • This endeavor might be challenging, given that clinical evidence that can guide the selection of biopsy size thresholds is limited
  • More importantly, in this large study, most nodules were categorized as TR3 or TR4, and overall the mean nodule size was 1.5 cm, representing a group of patients in whom management recommendations by the major RSSs are highly variable
  • These findings underscore the value of a care model that uses the best available evidence to understand thyroid cancer risk and supports the collaboration of patients and clinicians when deciding how to respond to this risk

#Arrangoiz #ThyroidSurgeon #CancerSurgeon #ThryoidExpert #ThyroidNodules #ThyroidCancer #Miami