Correlating thyroid cytology and histopathology: Implications for molecular testing.

Area of Science:

• Oncology
• Pathology
• Genomics

Background:

• Gene expression classifiers (GEC) are proposed for managing indeterminate thyroid nodules, potentially avoiding surgery.
• The study evaluates the negative predictive value (NPV) of a GEC in an academic setting.

Purpose of the Study:

• To assess the real-world negative predictive value (NPV) of a gene expression classifier (GEC) for indeterminate thyroid nodules.
• To determine if GEC performance meets acceptable standards for clinical decision-making.

Main Methods:

• Retrospective analysis of cytologically indeterminate fine-needle aspirates (FNAs) over a 3-year period.
• Histopathologic findings were correlated with GEC results.
• NPVs were calculated using published sensitivity and specificity data for the GEC.

Main Results:

• Out of 178 nodules, 71 (40%) were malignant.
• Malignancy prevalence varied by class: 41% (III), 29% (IV), and 96% (V).
• The overall GEC NPV was 91%, with class-specific NPVs of 88% (III), 92% (IV), and 26% (V).

Conclusions:

• The GEC's NPV in certain nodule classes, particularly class V, is significantly lower than desired.
• Clinical criteria alone may offer a more reliable NPV in some cases.
• Over-reliance on GEC without considering nodule class and prevalence may increase missed cancer diagnoses.