T. M. Cotton1, X. Jing1, S. John1, R. Lirov1, B. Miller1, M. Cohen1, G. Paul1, D. Hughes1 1University Of Michigan,Ann Arbor, MI, USA
Introduction:
Limitations of intraoperative frozen section (FS) for indeterminate thyroid lesions are well-appreciated. Studies about FS utility have involved ‘follicular lesions’ (FL) and were before the Bethesda System for Reporting Thyroid Cytopathology (BSRTC) subcategorized indeterminate FNA results into follicular lesions of undetermined significance (FLUS) and suspicious for follicular neoplasm (SfFN). We hypothesize that FS will have less utility with FLUS lesions compared to SfFN due to improved cytological categorization in the post-Bethesda era.
Methods:
From 2008 to 2014, 479 patients underwent thyroid lobectomy. Pre-Bethesda (2008-2011) patients with an FNA diagnosis of FL and post-Bethesda (2011-2014) patients with an FNA diagnosis of FLUS or SfFN who underwent FS were identified. The study group was comprised of 135 patients (65 FL, 45 FLUS, and 25 SfFN). The sensitivity, specificity, PPV, and NPV of FS within these categories was compared to final histopathology.
Results:
In the FL group, 6/65 patients were found to have thyroid cancer within the sampled nodule on histopathology with FS having a sensitivity of 50%, specificity of 100%, PPV of 100%, and NPV of 95%. Three of the six cancer cases were identified on FS and changed the operation in 3/65 FL patients (4.6%). In the FLUS group, 5/45 patients were found to have thyroid cancer within the sampled nodule on histopathology with FS having a sensitivity of 20%, specificity of 100%, PPV of 100%, and NPV of 91%. One of the five cancer cases was identified on FS and changed the operation in 1/45 FLUS patients (2.2%). In the SfFN group, 4/25 patients were found to have thyroid cancer within the sampled nodule on histopathology with FS having a sensitivity of 50%, specificity of 100%, PPV of 100%, and NPV of 91%. Two of 4 cancer cases were identified on FS and changed the operation in 2/25 SfFN patients (8%). No patients were over-treated due to false positive FS and 3/70 patients (4.3%) avoided reoperation across post-Bethesda groups.
Conclusion:
FS had limited value before BSRTC, but is even less useful in FLUS patients in the post-Bethesda era. FS had a higher sensitivity in SfFN than FLUS patients and avoided reoperation in 8% of SfFN patients compared to 2.2% (FLUS) and 4.6% (FL).