Effect of Processing and Fixation Time on IHC Results for Breast Biomarkers.
Stella Quan, William Pierceall, Michelle Wolfe, Lakshmi Alaparthi, Jeffery Kutok, Brian Ward. On-Q-ity, Waltham, MA; Brigham and Women's Hospital, Boston, MA
Background: IHC Biomarker expression levels of FFPE-derived sections have been reported to discriminate therapeutic response in a variety of solid tumor malignancies. While the field has generally employed subjective classification by trained pathologists using standard 1+, 2+, 3+ scoring strategies, advances in digital pathology and image analysis software offer the potential for more objective analysis using specific computer scoring algorithms. The goal of our development program is to create accurate cut-off values for prediction of therapeutic response. However, potential technical issues remain that may hinder correct placement of patients within a biomarker expression continuum. One potential obstacle may be the effect of variability in processing and fixation time on the protein expression assayed by IHC as scored by digital pathology systems.
Design: To investigate the magnitude of this potential obstacle, we collected breast resection tumor samples from 6 individual patients and divided these tumors into 6 aliquots. Each aliquot was immediately fixed in 10% neutral buffered formalin for 2, 8 or 48 hrs, or was held overnight after at 4oC in saline before fixing. Sections were stained with three biomarkers using optimized automated protocols. Tumor regions were annotated and scored with user defined macros. Variation in staining quantity and intensity, defined by a multiplicative formula for computer generated values, were compared between the various processing and fixation times.
Results: BRCA1, XPF and ATM nuclear localized IHC signal was assessed and the scores derived did vary with fixation time and temperature. Increasing fixation time from 2 to 48 hours enhanced the ability to detect antigen. The variation was not uniform across all markers as BRCA1 was shown to be the most variable with scores from digital pathology analysis within individual patients ranging from 0 to 275 (scale 0-300). While ATM and XPF showed less dramatic variation among fixation conditions, enhanced antigen detection was a consistent trend with the longer fixation of 48 hrs.
Conclusions: In contrast to previous studies on the expression of ER and PR, the DNA repair nuclear localized proteins examined in this study display substantial variation in macro-derived scores due to fixation. These data indicate that uniform fixation will eliminate one source of variation for IHC results, and is a necessary precursor to proper patient placement in studies designed to develop diagnostic algorithms to identify chemotherapy responders and non-responders with high sensitivity and specificity.
Monday, February 28, 2011 1:00 PM
Poster Session II # 270, Monday Afternoon