External Validation of a microRNA Thyroid Classifier: A Real-World Prospective Study.

Summary

In a prospective, independent, real-world validation of 256 Bethesda III/IV nodules, the microRNA classifier achieved 83% sensitivity, 83.5% specificity, and 93.6% NPV, supporting clinical decisions in >89% of cases and reducing surgeries by 79.5%. These data substantiate the utility of microRNA-based testing to manage indeterminate nodules and avoid unnecessary thyroidectomy.

Key Findings

• Prospective external validation in 256 indeterminate (Bethesda III/IV) nodules.
• Sensitivity 83.0%, specificity 83.5%, PPV 62.8%, NPV 93.6%.
• Clinical decision support in 95.5% (negative) and 89.8% (positive) cases; surgery reduction by 79.5%.

Clinical Implications

For Bethesda III/IV nodules, integrating this microRNA test into the diagnostic pathway can reduce diagnostic thyroidectomies, prioritizing surgery for test-positive patients while safely surveilling test-negative nodules with high NPV.

Limitations

• Limited surgical verification among test-negative cases may inflate NPV
• Single test platform; generalizability across laboratories and populations requires further study

Future Directions

Conduct multicenter comparative validations versus other molecular classifiers, assess cost-effectiveness, and establish longitudinal outcomes (malignancy rates in test-negative nodules under surveillance).