New Visualization Technique Enables Her2 Immunofluorescence Staining To Resemble Traditional Chromogenic DAB Stain
Mike Lazare, Kevin Kenny, Natalia Jun, Ainura Kyshtoobayeva, Denise Hollman, Dave Henderson, Alex Corwin, Tommy Ha, Sireesha Kaanumalle, Colin McCulloch, Kenneth Bloom. GE Global Research, Niskayuna, NY, Vanuatu; Clarient Diagnostics, a GE Healthcare Company, Aliso Viejo, CA
Background: The current clinical method for testing Her2 status involves using chromogenic immunohistochemistry (IHC) to detect Her2 protein expression. Chromogenic stains have several limitations that can be overcome by using immunofluorescence (IF). IF allows for the collection of linear, quantitative data across a greater dynamic range than chromogenic stains, which is particularly useful for Her2. However, conventional IF images can be difficult for pathologists to read compared to the commonly accepted, high contrast, chromogenic stains such as DAB. Here, we demonstrate the ability to convert traditional grayscale IF images into “molecular DAB” (mDAB) images which simulate the appearance of true DAB. This gives the pathologist the benefits of IF while making the images as readable as a chromogenic stain.
Design: IF and DAB staining was performed on 12 FFPE invasive ductal breast carcinoma samples that exhibited varying degrees of Her2 expression, collected between June 2011 and March 2012. A mAb for Her2 was used for conventional DAB staining. The same antibody was conjugated to Cy5 dye for IF staining on serial sections. Two board-certified pathologists scored both the mDAB and DAB images on a 0/1+/2+/3+ scale. In each patient where mDAB & DAB scores were discordant, alterations in Herceptin treatment recommendations were recorded in light of the patient's clinical Her2/CEP17 amplification ratios via FISH.
Results: Pathologists 1 and 2 had consistent scoring between mDAB and DAB in 8/12 cases (kappa=0.29) and 10/12 cases (kappa=0.68), respectively. The two pathologists agreed in 12/12 mDAB images (kappa=1) but only 8/12 DAB images (kappa=0.29). Of the 5 cases of discordance between mDAB & DAB for either pathologist, 3 would have been given the same Herceptin recommendation based on the mDAB or the DAB scoring. In one case the mDAB score would have led to Herceptin treatment of a case with confirmed Her2 gene amplification via FISH, while the DAB scores would not have resulted in Herceptin treatment. FISH data was not available in the last discordant patient. It is noteworthy that there were no discordant cases among the mDAB reads.
Conclusions: Immunofluorescence staining offers many technical advantages over traditional brightfield immunohistochemistry. Despite these advantages, interpretation of raw darkfield images can be challenging. Therefore, this molecular DAB technique offers a promising method for rendering IF images in a manner more familiar to pathologists; therefore combining the best of both worlds.
Wednesday, March 6, 2013 9:30 AM
Poster Session V # 266, Wednesday Morning