1173P - Combining mass spectrometry with quantitative continuous scoring to unlock the full quantitative potential of immunohistochemistry

Background

Accurately measuring drug target expression levels is critical for understanding drug efficacy and advancing biomarker discovery. Target expression is often measured by immunohistochemistry (IHC) using visual pathology assessment. Digital pathology-based Quantitative Continuous Scoring (QCS) has emerged as a more precise alternative. This study aims to investigate the correlation between target protein levels measured by mass spectrometry (MS) and IHC visual or digital scores; as well as to evaluate if this orthogonal analysis can guide the development of fit-for-purpose IHC assays.

Methods

PDX FFPE samples (n=103) were characterized by targeted MS and stained with 7 research use only IHC assays. IHC images were quantified using visual H-scores (0-300) and by QCS. Quantitative fluorescence-activated cell sorting (qFACS)-derived antibody-binding capacity for one of the targets was quantified in 12 cell lines, which were also characterized by MS and IHC.

Results

For all analyzed targets, both H-scores and QCS correlated with the MS protein concentrations following a sigmoid function from which the assay’s dynamic range could be estimated. Consistently across targets, visual scores reached saturation (H-scores >270) at lower protein concentration than QCS, suggesting that the dynamic range of an IHC assay can be expanded using digital-scoring methods. Given the correlation between IHC scores and protein concentrations, we next tested whether the number of proteins per-cell could be estimated directly from IHC images for one of the examined targets. Using a conversion formula derived from cell lines characterized by MS, qFACS and IHC, we determined that the average number of proteins-per-cell could be estimated from both IHC visual and digital readouts.

Conclusions

QCS more precisely quantifies IHC target expression levels and expands the IHC scoring dynamic range as compared to visual pathology assessment. The improved IHC quantification unlocks the possibility of using MS to guide IHC assay development, including optimizing its range.

Clinical trial identification

Editorial acknowledgement

Legal entity responsible for the study

AstraZeneca.

Funding

AstraZeneca.

Disclosure

A. Hidalgo-Sastre, J. Zaucha, N. Damond, T.H. Tan, A. Meier, M. Schick, M. Korade, A. Kunihiro, A. Chambers, N. Durham, H. Sade, Y.J. Kim, M. Gustavson, M.C. Rebelatto, N. Giraldo: Financial Interests, Institutional, Full or part-time Employment: AstraZeneca. J.C. Barrett: Financial Interests, Institutional, Full or part-time Employment: Precede Biosciences. G. Schmidt: Financial Interests, Personal, Full or part-time Employment: AstraZeneca; Financial Interests, Personal, Stocks or ownership: AstraZeneca. J.S. Reis-Filho: Financial Interests, Institutional, Full or part-time Employment: AstraZeneca; Financial Interests, Personal, Leadership Role: Grupo Oncoclinicas; Financial Interests, Personal, Advisory Board: Repare Therapeutics, PaigeAI, Genentech/Roche, Invicro, Ventana Medical Systems, Volition RX, PaigeAI, Goldman Sachs, Novartis, Repare Therapeutics, Personalis, SAGA Diagnostics, Bain Capital Life Sciences; Financial Interests, Personal, Stocks or ownership: Repare Therapeutics, PaigeAI.