|
Activity Number:
|
25
- Medical Devices and Diagnostics Speed Session
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Sunday, August 8, 2021 : 1:30 PM to 3:20 PM
|
|
Sponsor:
|
Section on Medical Devices and Diagnostics
|
|
Abstract #318729
|
|
|
Title:
|
Quantile Index Protein Biomarkers Based on Cell-Level Immunohistochemistry Data
|
|
Author(s):
|
Misung Yi* and Inna Chervoneva
|
|
Companies:
|
Department of Pharmacology and Experimental Therapeutics, Thomas Jefferson University and Department of Pharmacology and Experimental Therapeutics, Thomas Jefferson University
|
|
Keywords:
|
Functional regression quantile index;
Functional Cox model;
Tissue microarrays;
Immunohistochemistry biomarkers;
Breast cancer
|
|
Abstract:
|
Protein biomarkers of cancer progression and response to therapy are increasingly important for advancing personalized medicine. The data motivating this work includes protein expression levels in cancer tissue microarray from 1,000+ breast cancer patients. Advanced quantitative pathology platforms enable measurement of proteins in individual cells, but only mean expression within the region of interest is usually considered as a biomarker. In this work, we consider the entire empirical quantile function Q(p) computed for a sample of cellular signal intensity levels in a tissue as a predictor of clinical outcome. The proposed Functional Regression Quantile Index (FR-QI) protein biomarker is defined as the integral of Q(p) multiplied by common weight function ?(p), where ?(p) is an unknown function represented by a penalized spline. Since the primary outcome of interest is the progression free survival (PFS), we use the functional Cox model framework to optimize FR-QI as a predictor of PFS. The new approach yields novel biomarkers for proteins that do not have a prognostic value if only the mean signal intensity is considered.
|
Authors who are presenting talks have a * after their name.
