Use of Image Analysis in Immunohistochemistry (IHC) Assay Development and Quality Control.
Scott Webster, Charlotte Roach, Vaishali Tanna, Julie Erickson, Karen Wakamiya. Dako North America, Inc., Carpinteria, CA
Background: Image analysis is gaining increasing acceptance as a tool for the interpretation of IHC stains in the pathology laboratory. A number of slide scanners are now available that provide algorithms designed for the assessment of clinical markers such as HER2, ER/PR and Ki-67. In addition, such platforms can have utility beyond their primary duty of evaluating clinical specimens. For example, image analysis can provide benefit to activities such as assay development and daily quality control, thereby adding value across the spectrum of laboratory workflow.
Design: Activities typically performed in the development of IHC assays include titering the primary antibody, defining scoring approaches and establishing quality control methods. To illustrate the application of image analysis to such activities, an assay for ßIII-Tubulin (ß3T) IHC currently in development in our laboratory was used as an example. Stained slides were scanned using the ACIS® III and scored using an algorithm based on the intensity of the IHC stain. Where required, slides were manually scored in parallel using the H-score method (Budwit-Novotny et al., 1986, Cancer Res 46:5419-5425).
Results: Antibody titration. Intensities from slides stained using sequential titrations of different anti-ß3T batches were plotted against dilution factor. The resulting plots were used to calculate the dilutions required to obtain equivalent staining. These dilutions compared favorably to the IgG concentrations of the antibody batches (p<0.05). Scoring IHC stains. Image analysis results exhibited a significant correlation with manual assessment (H-score), demonstrating the ability to replicate complex manual scoring schema (r=0.79, p<0.05). Reproducibility/Quality control. As a model for tracking staining batch controls, tissue sections from 27 consecutive experiments (stained as part of precision studies) were scanned, their IHC intensities plotted in a Levey-Jennings style chart and evaluated using Westgard rules:
Conclusions: These results demonstrate the successful application of image analysis methods to laboratory activities above and beyond the primary use of assessing clinical specimens. This suggests additional value for the technology, particularly in laboratories that engage in development and/or QC activities.
Monday, February 28, 2011 1:00 PM
Poster Session II # 277, Monday Afternoon