Age-specific risk of malignancy in pediatric thyroid cytology: Reframing ROM based on pre-test probability

Area of Science:

• Endocrinology
• Paediatric Oncology
• Cytopathology

Background:

• Paediatric thyroid nodules, while rare, exhibit a higher malignancy rate compared to adult nodules.
• Current Bethesda risk-of-malignancy (ROM) estimates lack age stratification and are susceptible to verification bias.

Purpose of the Study:

• To develop age-specific ROM, likelihood ratios (LRs), and post-test malignancy probabilities for paediatric and young adult thyroid cytology.
• To address limitations of existing ROM estimates by incorporating age and quantifying verification bias.

Main Methods:

• Analysis of 2,728 thyroid fine-needle aspirations (FNAs) from patients aged 0-25 years across multiple tertiary centers (2000-2023).
• Classification using the 2023 Bethesda System and grouping into four age bands (0-8, 9-14, 15-18, 19-25 years).
• Calculation of lower-bound and surgery-only upper-bound ROM, with post-test probabilities computed using Bayes' theorem and age-specific pretest probabilities.

Main Results:

• Malignancy prevalence in operated nodules decreased with age, from 84.2% (0-8 years) to 64.6% (19-25 years).
• Verification bias was significant, with surgery-only ROM (67.7%) substantially higher than lower-bound ROM (24.6%).
• Indeterminate cytology categories showed age-related variations in post-test risk, e.g., follicular neoplasm ~71% risk in a 9-year-old vs. ~49% in a 24-year-old.

Conclusions:

• Age is a critical factor influencing both pretest and post-test malignancy probabilities in paediatric thyroid cytology.
• An age-stratified LR framework effectively quantifies verification bias and provides individualized risk estimates.
• These age-specific estimates are crucial for informed decision-making regarding surgery, molecular testing, and follow-up in young patients.