Abstract
Abstract
Introduction/Objective
Accurate correlation between intraoperative frozen section (FS) and subsequent permanent section (PS) diagnoses is essential for quality patient care. Intraoperative FS is used for confirmation of diagnoses and margin status (MS) evaluation in head and neck (H&N) pathology. Reported FS-PS discrepancy rates for H&N cases range between 1.42% to 1.80%, mostly due to sampling and interpretation errors. We sought to review our experience with FS and concordant PS from H&N pathology cases diagnosed at a high volume, university medical center.
Methods/Case Report
Correlation was performed between intraoperative FS and paraffin-embedded PS of H&N surgical resections between January 2019 to December 2021. FS-PS discrepancies were classified due to “sampling” or “interpretation”. Recurrence rates based on a 2 year follow up period of 2019 cases was also recorded.
Results (if a Case Study enter NA)
2337 H&N cases yielded 11108 intraoperative FS diagnoses. FS-PS correlation identified 142 (1.27%) discrepancies. Of these, 85 (60%) were for MS control, comprised of 45 (53%) due to “sampling” and 40 (47%) to “interpretation”. FS slides generally represented a single top section taken from the tissue, with “sampling” discrepancies due to discordant diagnoses identified on a deeper cut of tissue on the PS slides.
Common “interpretative” discrepancies included overlooking high-grade dysplasia. For patients diagnosed in 2019, we found that 6% of discrepant cases of locoregional disease over a 2 year follow up period were due to inadequate margin control.
Conclusion
Our institutional intraoperative FS-PS discrepancy rate for H&N pathology cases is 1.27%, and within the national quality standards. Our FS-PS discrepancy rate for H&N cases related to MS control were mostly due to “sampling". Intraoperative review of deeper cuts of FS tissue and in service instruction for pathologists regarding intraoperative reporting of high-grade dysplasia will be the subject of a prospective study for potential improvement of “sampling” and “interpretative” discrepancies, respectively. Future studies hope to minimize the cases of locoregional occurence due to inadequate margin status.