Multiplatform comparison of molecular oncology tests performed on cytology specimens and formalin-fixed, paraffin-embedded tissue

Abstract

BACKGROUND Molecular oncology testing is important for patient management, and requests for the molecular analysis of cytology specimens are increasingly being made. Formalin-fixed, paraffin-embedded (FFPE) cell blocks of such specimens have been routinely used for molecular diagnosis. However, the inability to assess cellularity before cell block preparation is a pitfall of their use. In this study, various cytologic preparations were tested with several molecular test platforms, and the results were compared with paired FFPE tissue. METHODS Seventy-seven cytology cases, including fine-needle aspiration smears, touch preparations, and SurePath thin-layer preparations, were selected from the archives. Areas of interest were removed from the slide with a matrix capture solution. DNA extracted from the cells was evaluated by mutation analysis for BRAF, epidermal growth factor receptor (EGFR), RAS, and a 50-gene panel with various testing platforms (single-nucleotide primer extension assay, Sanger sequencing, and next-generation sequencing). Thirty-eight tumors with available FFPE tissue were tested in parallel. RESULTS The average DNA concentration was 299 ng/μL for the cytology specimens and 171 ng/μg for the paired FFPE tissue. Point mutations and large deletions were detected in BRAF, KRAS, NRAS, HRAS, and EGFR genes. In comparison with FFPE tissue, 5- to 8-fold less input DNA was needed when cytology preparations were used. The concordance between cytology specimens and FFPE tissue was 100%. CONCLUSIONS Cytologic preparations were found to be a reliable source for molecular oncology testing. DNA derived from cytology specimens performed well on multiple platforms, and 100% concordance was observed between cytology specimens and FFPE tissue.