A targeted proteomics assay for the preoperative diagnosis of thyroid nodules.
Summary
ThyroProt, combining a 3-protein targeted proteomic signature with BRAF V600E and demographics, achieved AUC 0.94 and 90.7% accuracy in a prospective multicenter test set, with 100% specificity in Bethesda III/IV nodules. Consistent performance across independent cohorts supports clinical adoption to reduce unnecessary diagnostic surgery.
Key Findings
- Prospective, blinded, multicenter validation: AUC 0.94 and 90.7% accuracy in 322 test FNAs.
- In Bethesda III/IV nodules, sensitivity 82.4% and specificity 100%, accuracy 88.0%.
- External multicenter cohorts maintained AUC 0.87–0.91 and accuracy 84.3%–85.7%.
Clinical Implications
Supports integrating targeted proteomics with molecular testing in FNA workflows to triage indeterminate nodules, potentially decreasing diagnostic lobectomies and expediting appropriate management.
Why It Matters
Provides a scalable, validated diagnostic that addresses a major gap in thyroid nodule management—indeterminate cytology. Prospective, blinded, multicenter validation and external replication bolster immediate translational value.
Limitations
- Health-economic impact and real-world workflow integration were not assessed.
- Access to targeted mass-spectrometry platforms may vary across settings.
Future Directions
Prospective impact studies on surgical decision-making, cost-effectiveness analyses, and harmonization of proteomics platforms to support broad clinical deployment.
Study Information
- Study Type
- Cohort
- Research Domain
- Diagnosis
- Evidence Level
- II - Prospective diagnostic accuracy study with external validation
- Study Design
- OTHER