Microdissection and molecular genetic analysis of HER2/neu in breast carcinoma.
Academic Article
Overview
abstract
Precise correlation of histomorphology with molecular genetic analysis is difficult in tissues composed of heterogeneous cell populations. We describe here a novel microdissection technique employed to correlate HER2/neu (HER2) immunohistochemical staining with HER2 genetic analysis in formalin-fixed, paraffin-embedded breast tissue. Fourteen invasive ductal carcinomas were selected from the pathology files of Memorial Sloan-Kettering Cancer Center that had been immunostained for HER2. Seven tumors showed typical membrane immunoreactivity and seven were negative. A dissecting microscope was then used to isolate minute (< or = 1 mm x 1 mm) areas of invasive carcinoma and normal breast tissue for molecular study. To document the type of cell sample submitted for polymerase chain reaction (PCR) analysis, each microdissected piece of tissue was photographed prior to removal from the glass slide. A preliminary study of four cases compared the results of PCR and genetic analysis using microdissected hematoxylin and eosin (H & E)-stained tissue, unstained dewaxed tissue, and destained dewaxed tissue in four specimens. Similar results were obtained with all three tissue preparations. Thereafter, H & E stained sections were selected as the tissue preparation of choice because tissue details were seen more clearly. There was complete correlation of immunohistochemical staining and HER2 analysis by PCR in all 14 cases. In the final 10 cases, the PCR product was resolved by gel electrophoresis and quantified by optical densitometry. Fourfold to eightfold amplification of HER2 was found in the five tumor specimens that immunohistochemically stained for HER2. A single copy of HER2 was found in all HER2-negative tumors and in normal breast tissue. We conclude that it is possible to quantify gene amplification of HER2 in minute samples of H & E-stained normal and malignant breast tissue. This microdissection technique can be applied to correlative histologic--molecular genetic analysis in a wide variety of tumor types.
