Abstract 153: Development of a workflow to handle the quality control and analysis of Olink protein biomarker data in early phase oncology clinical trials

Abstract

Abstract INTRODUCTION: Olink is a high-throughput targeted biomarker platform based on Proximity Extension Assay (PEA) technology using oligonucleotide-labeled antibody pairs. At AbbVie RWC, we have primarily utilized the 96-plex Immune-Oncology (IO) panel measuring 92 proteins to support our early-phase oncology clinical trials. As we analyzed batches of specimens from ongoing trials, we set out to develop a workflow to control data quality and streamline analysis. METHODS: Prior to sample shipment to Olink, we randomize sample position onto plates to ensure key factors such as dose and cohort are balanced within and across plates. This prevents plate or location effects from becoming a confounder. If the study specimens are analyzed in more than one batch, subsequent batches require bridging samples from the first batch to allow us to detect and remove batch effects if necessary. Our workflow thus identifies bridging samples and recommends the optimal number. Upon receiving the data, we evaluate its quality by calculating inter- and intra-plate coefficients of variation (CVs) and visualizing it in multiple ways. The next step in our workflow is analysis, including pharmacodynamic (PD) analysis and association with clinical response. In the PD analysis, we identify biomarkers that significantly change across time and/or exhibit a dose effect. To help interpret the results, enrichment analysis is performed to identify biological processes enriched in the significantly modulated biomarkers. Heatmaps and volcano plots are also generated. For analysis of association with clinical response, we examine how each biomarkers' baseline value and modulation over time associates with progression free survival (PFS) and best overall response (BOR). RESULTS: We successfully created a workflow that handles quality control and analysis of Olink protein biomarker data. The workflow starts from experimental design quality control. Upon receiving the data, we conduct data quality assessment and examination and removal of potential batch effects. Finally, we perform analysis of the data to identify potential biomarkers of interest. We used elements of this workflow on data from M15-891 (ABBV-181) and have now applied the full workflow to multiple batches of data from three early-phase oncology trials (M15-862 (ABBV-927), M16-074 (ABBV-368) and M19-345 (ABBV-151)), with plans to incorporate this methodology into several additional upcoming trials. CONCLUSIONS: We have developed a comprehensive workflow that manages Olink quality control and analysis and can potentially be implemented to all early-phase oncology clinical trials using Olink. It allows us to standardize and streamline Olink data handling and provide robust and efficient analyses that enable rapid biomarker data interpretation, hypothesis generation, and decision making for our trials. Citation Format: Claire J. Guo, Mary Saltarelli, Stacie Lambert, Hua Fang, Chun Zhang. Development of a workflow to handle the quality control and analysis of Olink protein biomarker data in early phase oncology clinical trials [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 153.