Examination of Independent Prognostic Power of Gene Expressions and Histopathological Imaging Features in Cancer

Abstract

Cancer prognosis is of essential interest, and extensive research has been conducted searching for biomarkers with prognostic power. Recent studies have shown that both omics profiles and histopathological imaging features have prognostic power. There are also studies exploring integrating the two types of measurements for prognosis modeling. However, there is a lack of study rigorously examining whether omics measurements have independent prognostic power conditional on histopathological imaging features, and vice versa. In this article, we adopt a rigorous statistical testing framework and test whether an individual gene expression measurement can improve prognosis modeling conditional on high-dimensional imaging features, and a parallel analysis is conducted reversing the roles of gene expressions and imaging features. In the analysis of The Cancer Genome Atlas (TCGA) lung adenocarcinoma and liver hepatocellular carcinoma data, it is found that multiple individual genes, conditional on imaging features, can lead to significant improvement in prognosis modeling; however, individual imaging features, conditional on gene expressions, only offer limited prognostic power. Being among the first to examine the independent prognostic power, this study may assist better understanding the "connectedness" between omics profiles and histopathological imaging features and provide important insights for data integration in cancer modeling.

Keywords: cancer prognosis; histopathological imaging features; independent prognostic power; omics profiles.

Figure 1

Flowchart of extracting imaging features. Step 1: whole-slide histopathology images are cropped into small subimages of 500 × 500 pixels, and 20 subimages are then randomly selected. Step 2: Imaging features are extracted using CellProfiler [19] for each subimage. Step 3: For each patient, features are averaged.

Figure 2

Gene ontology (GO) and pathway enrichment analysis of the identified genes. (a) lung adenocarcinoma (LUAD), (b) liver hepatocellular carcinoma (LIHC).

Figure 3

Kaplan–Meier (KM) curves for low (blue) and high (red) risk groups under models A1 and A2. (a,b) for LUAD: Gene HNRNPK as well as selected imaging features (A1), and only selected imaging features (A2). (c,d) for LIHC: Gene GOT2 as well as selected imaging features (A1), and only selected imaging features (A2). p values are computed from log rank tests.

Figure 4

KM curves for low (blue) and high (red) risk groups under models B1 and B2. (a,b) for LUAD: Imaging feature Texture_Correlation_maskosingray_3_00 as well as selected gene expressions (B1), and only selected gene expressions (B2). (c,d) for LIHC: Imaging feature StDev_Identifyhemasub2_Texture_DifferenceEntropy_ImageAfterMath_3_00 as well as selected gene expressions (B1), and only selected gene expressions (B2). p values are computed from log rank tests.