Date of Award

January 2024

Document Type

Open Access Thesis

Degree Name

Medical Doctor (MD)

Department

Medicine

First Advisor

Luisa F. Escobar-Hoyos

Second Advisor

Kenneth R. Shroyer

Abstract

Background: In an era where tumor heterogeneity leads to widely variable cancer treatment outcomes, there is an unmet need for biomarkers that can characterize tumor subtypes, inform prognosis, and guide clinical decision-making for therapy. As an oncofetal monofilament that is only expressed in embryonic and cancer tissue, keratin 17 (K17) has been demonstrated as an effective biomarker in a wide variety of cancer types, detection methods, and clinical purposes. In order to guide and inform future investigations on K17, we performed a systematic review of clinical studies assessing the effectiveness of K17 as a cancer biomarker.

Methods: We performed a literature search of relevant articles reporting K17 as a cancer biomarker prior to December 16th, 2023. After manual review, we assigned articles to diagnostic, prognostic, and predictive study categories. We extracted relevant diagnostic and prognostic statistics, experimental methodologies, and co-examined biomarkers. Additionally, we performed bias analysis utilizing data reported from prognostic studies. Finally, using RNA-seq samples of tumor and normal tissue specimens from The Cancer Genome Atlas (TCGA) and Genotype-Tissue Expression (GTEx) portal, we perform our own diagnostic analysis across 21 tumor/normal tissue pairs, as well as Kaplan-Meier analysis across 32 cancer types.

Results: Of the 453 studies identified in our literature search, we obtained 40 diagnostic, 50 prognostic, and 5 predictive studies. Altogether, we summarized the methodologies of 37 diagnostic studies and ROC (Receiver Operating Curve) analyses provided by 15 articles. Our review suggests that K17 is a highly applicable diagnostic biomarker, characterizable by multiple assay types, to distinguish either malignant from normal pathology, or two different malignant pathologies. For prognostic studies, we extracted Cox proportional hazards models provided by 36 studies, a majority of which suggest that K17 is a negatively prognostic biomarker in a wide range of cancer types, including oropharyngeal, esophageal, gastric, pancreatic, gallbladder, ovarian, and endometrial cancers. Finally, based on our review, K17 is predictive of inferior response to therapy in 4 out of 5 predictive studies.

Conclusion: K17 is a widely applicable cancer biomarker that has been shown to be effective in diagnostic settings, and increasingly in patient specimens obtained non-invasively. As both an IHC and RNA-based biomarker, K17 is widely associated with negative prognosis. New predictive studies are beginning to correlate K17 to inferior chemotherapeutic and immunotherapeutic treatment response.

Comments

This is an Open Access Thesis.

Download
Open Access

This Article is Open Access

Share

COinS