Limitations of Explainability for Established Prognostic Biomarkers of Prostate Cancer.

Kalifa Manjang,Olli Yli-Harja,Frank Emmert-Streib,Matthias Dehmer

doi:10.3389/fgene.2021.649429

Abstract

High-throughput technologies do not only provide novel means for basic biological research but also for clinical applications in hospitals. For instance, the usage of gene expression profiles as prognostic biomarkers for predicting cancer progression has found widespread interest. Aside from predicting the progression of patients, it is generally believed that such prognostic biomarkers also provide valuable information about disease mechanisms and the underlying molecular processes that are causal for a disorder. However, the latter assumption has been challenged. In this paper, we study this problem for prostate cancer. Specifically, we investigate a large number of previously published prognostic signatures of prostate cancer based on gene expression profiles and show that none of these can provide unique information about the underlying disease etiology of prostate cancer. Hence, our analysis reveals that none of the studied signatures has a sensible biological meaning. Overall, this shows that all studied prognostic signatures are merely black-box models allowing sensible predictions of prostate cancer outcome but are not capable of providing causal explanations to enhance the understanding of prostate cancer.

Highlights

Prostate cancer (PCa) is the second most prevalent cancer among men, the average age of diagnosis is 66 years, and about 60% of diagnosed cases occur in men over 65 years old
In order to investigate this, we study 32 published prognostic PCa signatures from the literature and demonstrate that random gene sets can be found with similar prediction capabilities as these signatures but opposite biological meaning
Size of Biomarker Sets and gene ontology (GO)-Terms in Signatures In Table 1, we show an alphabetically ordered overview of all 32 prognostic BM signatures included in our analysis

Summary

Introduction

Prostate cancer (PCa) is the second most prevalent cancer among men, the average age of diagnosis is 66 years, and about 60% of diagnosed cases occur in men over 65 years old. In the United States, for example, 191, 930 newly diagnosis cases of PCa are estimated in 2020, resulting in about 33, 330 mortalities (Siegel et al, 2020). A substantial proportion of PCa is characterized as slow-growing and indolent requiring no immediate therapeutic intervention. Tumor stages T1 and T2, and tumor stages higher than T2 are more aggressive and invade the surrounding organs and the patient is more likely to die from the disease (Chen et al, 2020). For men with local or regional PCa, the 5-year survival rate is almost 100%, whereas the 5-year survival rate for men with metastatic PCa is 31%.

Objectives

Methods

Results

Discussion

Conclusion