Please select Print from the file menu or right mouse click and choose print from the menu to print this page.

[1594] Quantitative Fluorescence-Based Analysis of Multiplexing for Immunohistochemistry

M Cregger, M Harigopal, S Siddiqui, RL Camp, DL Rimm. Yale University School of Medicine, New Haven, CT

Background: Immunohistochemistry (IHC) allows assessment of protein expression levels without loss of spatial context. Recently, a number of groups have begun multiplexing IHC to assess multiple markers on the same slide to maximize use of scare tissue resources and/or to optimally assess the co-localization of multiple protein targets. However, multiplexing can be interpreted in many ways and has rarely been validated. Here we define 2 potential methods of multiplexing and validate each using AQUA, a new method of in situ quantitative IHC.
Design: The common methods of multiplexing include 1) compartmental multiplexing (where antibodies are mixed and intensity is assessed on the basis of their subcellular or architectural localization), and 2) true multiplexing using multiple visualization tags on the same section. Historically multiplexing has been done with multiple enzymatic reactions (peroxidase and phophatase). Here we use only a single amplification system (peroxidase) for both methods, with intermediate HRP quenching. Each method is evaluated using a breast cancer tissue microarray with a selected set of cases and cell lines designed to optimize evaluation of estrogen receptor (ER).
Results: Her2 and ER were evaluated on a breast cancer patient cohort using both true multiplexing, and compartmental multiplexing. Linear regressions were performed to compare ER as a single marker versus true multiplexed ER showed high correlation (R=0.86). In contrast, compartmental multiplexing showed a lower correlation (R=0.66) and a scatterplot showing evidence of falsely elevated Her2 staining, presumably due to ER expression outside of the nucleus, contaminating the assessment of non-nuclear Her2. PR and p53 were used to assess the accuracy of simultaneous localization. The comparison of the multiplexed TMA with the single marker TMA shows R values for PR and p53 of 0.75 and 0.85 respectively. However, the correlation between PR and P53 is only 0.15 showing that the high correlation is not a spurious result due to common co-expression of both PR and p53.
Conclusions: While compartment multiplexing is conceivable, we show that it is inaccurate in the case of Her2 and ER. We show that that assessment in the same compartment, with HRP quenching between steps, can provide accuracy in multiplexed analysis that is comparable to that obtained when doing markers one at a time. We propose these methods as a standard for validation of multiplexing as new studies are done to optimize use of scarce tissue resources.
Category: Techniques

Monday, March 26, 2007 2:45 PM

Platform Session: Section G, Monday Afternoon