Optimization and validation of immunocytochemical detection of oestrogen receptors on cytospins prepared from fine needle aspiration (FNA) samples of breast cancer.

Objective: To optimize and validate immunocytochemical (ICC) assessment of oestrogen receptors (ERs) on cytospins prepared from fine needle aspiration (FNA) samples.

Methods: Optimal conditions and variability in ICC detection of ERs were established on cytospins prepared from the human breast cancer cell line MCF-7. Protocols that yielded adequate results were further validated on 52 FNA samples of resected breast cancer tumours using analysis of concordance with the ER status, determined by standard immunohistochemistry on corresponding formalin-fixed, paraffin-embedded tissue (FFPET). On 37 diagnostic FNA samples, manual immunostaining with antibody 1D5 was compared with automated immunostaining with antibody 6F11.

Results: The highest percentage of ER-positive MCF-7 cells with lowest variability was obtained on methanol-fixed cytospins with or without microwave pre-treatment: 72 ± 5% and 75 ± 7%, respectively. Microwave pre-treatment was mandatory for Papanicolaou-stained cytospins in order to achieve between 63 ± 14% and 67 ± 9% of ER-positive MCF-7 cells. The concordance between ICC assessment of ERs on FNA samples and corresponding FFPET sections was complete for methanol-fixed cytospins (100%, kappa = 1) and adequate for Papanicolaou-stained cytospins (94%, kappa = 0.84) and Papanicolaou-stained smears (92%, kappa = 0.75). Complete agreement in ICC detection of ERs was obtained for manual immunostaining with antibody 1D5 and automated immunostaining with antibody 6F11.

Conclusions: Methanol-fixed cytospins prepared from FNA samples ensure highly reliable ICC assessment of ERs, whereas Papanicolaou-stained cytospins or smears are conditionally suitable because of the small risk of false negative results.