Skip to main content

Frozen section analysis of margins for head and neck tumor resections: reduction of sampling errors with a third histologic level.

Olson SM, Hussaini M, Lewis JS Jr

Publication: Mod Pathol (2011) • Vol 24 • Page 665-70PMID: 21217647View on PubMed

Purpose

The purpose of this paper was to determine the effects of analyzing 3 rather than 2 frozen section margins intra-operatively to determine whether complete resection of tumor was achieved.

Conclusion

When specimens for which no tumor or dysplasia was identified on the frozen sections or the formalin-fixed paraffin-embedded (FFPE) section were excluded from analysis, overall and sampling error rates for two-level sectioning were significantly higher than three-level sectioning error rates, but interpretation error rates were not different. A significantly higher percentage of margins were found to harbor tumor in the frozen sections during the period when 3 sections were analyzed compared to during the period when 2 sections were analyzed. Keratinizing squamous cell carcinoma was the only tumor subtype examined that showed a significant increase in positive margin sampling error rates between two-level sectioning and three-level sectioning.

Studies (1)

Purpose

The purpose of this study was to determine the effects of analyzing 3 rather than 2 frozen section margins intra-operatively to determine whether complete resection of tumor was achieved.

Summary of Findings

When two frozen section margins were analyzed intra-operatively, the sampling and interpretation error rates were 1.6% and 0.8%, respectively for a combined error rate of 2.4%. When an additional third frozen section margin from a deeper level was examined, the sampling and interpretation error rates were 1.2% and 1.3%, respectively (2.5% overall), and were not significantly different from the two-level sectioning analysis. However, when specimens for which no tumor or dysplasia was identified on the frozen sections or an FFPE section were excluded, two-level sectioning error rates were significantly higher than three-level sectioning error rates, for both overall rates (22.9% versus 15.2%, p=0.03) and sampling error rates (15.3% versus 7.4%, p=0.006), but not for interpretation error rates (7.6% versus 7.8%, p=1.0). A significantly higher percentage of margins were found to harbor tumor in the frozen sections during the period when 3 section were analyzed (16.7%) compared to during the period when 2 sections were analyzed (10.5%, p<0.0001). Keratinizing squamous cell carcinoma was the only tumor subtype examined that showed a significant increase in positive margin sampling error rates between two-level sectioning and three-level sectioning.

  • Biospecimens

    TissueHead and Neck
  • Preservation Types

    FrozenFormalin
  • Diagnoses

    NeoplasticCarcinoma
    NeoplasticBenign
  • Platforms

    MorphologyLight microscopy
  • Pre-analytical Factors

    ClassificationPre-Analytical FactorValue(s)
    PreaquisitionDiagnosis/ patient condition
    Keratinizing squamous cell carcinoma
    Non-keratinizing squamous cell carcinoma
    Other tumors
    Biospecimen Aliquots and ComponentsAliquot size/volume
    2 frozen sections and 1 paraffin section examined
    3 frozen sections and 1 paraffin section examined