Abstract B066: Keratin 17 and GATA6 expression in pancreatic ductal adenocarcinoma: A subtyping stratification approach

Michael Horowitz,Lyanne Oblein,Jaymie Oentoro,Anika Porwal,Wei Jiang,Natalia Marchenko,Luisa Escobar-Hoyos,Kenneth Shroyer
DOI: https://doi.org/10.1158/1538-7445.panca2023-b066
IF: 11.2
2024-01-18
Cancer Research
Abstract:Background: Clinically and histologically identical pancreatic ductal adenocarcinomas (PDACs) differ in response to treatment, disease progression, and survival. Molecular subtyping defines patient populations that differ in terms of prognosis and response to gemcitabine and 5-FU-based chemotherapy. Our goal has been to develop biomarker-based Immunohistochemical protocols that are more accurate than molecular subtyping to guide individualized therapeutic intervention. We focused on two biomarkers highly associated with the basal and classical subtypes, respectively keratin 17 (K17), which promotes tumor growth and chemoresistance, and GATA6, a transcription factor involved in normal pancreatic development that is correlated with longer-term survival in PDAC. Methods: Formalin-fixed, paraffin-embedded (FFPE) tissue specimens (n=120) were selected from the archival collections of Stony Brook University Hospital and Thomas Jefferson University Hospital. Dual localization of K17 and GATA6 was performed by indirect immunoperoxidase methods utilizing Vina Green chromogen for K17 and diaminobenzidine (DAB) for GATA6. The detection of K17 staining in ≥ 10% of tumor cells was scored as high for K17 while the detection of moderate to low intensity staining for GATA6 was scored as low, as described (K. Duan et al. Sci Repts, 2021, 14951). Results: The analysis of IHC-stained PDAC revealed that K17 and GATA6 were heterogeneously expressed, with the strongest GATA6 expression in glandular components and the strongest K17 expression in single cells that diffusely infiltrated the stroma and peripancreatic soft tissue. Furthermore, K17 and GATA6 expression were not mutually exclusive (>50% of cases had high expression of both markers). Correlation of IHC test results in the context of tissue histomorphology revealed that PDAC is highly plastic, with the capacity to transition from well-differentiated to poorly differentiated states that are highlighted by changes in K17 and GATA6 expression. Furthermore, testing for both markers provided greater accuracy in prognostic stratification than testing either marker alone (K17: HR=1.670, p=0.0241; GATA6: HR=0.6755, p=0.0491; K17+GATA6: HR=2.008, p=0.0092). Additionally, high K17 expression, correlated with chemoresistance to gemcitabine, in both low and high-GATA6 cases, indicating that K17 has a dominant negative effect on treatment response. Conclusions: K17 and GATA6 are heterogeneously expressed but not mutually exclusive and highlight the plasticity of diffusely infiltrative versus glandular components of PDAC. Dual color IHC to score both markers provides enhanced prognostic accuracy compared to independent testing for K17 and GATA6. Moreover, high K17 expression overrides GATA6 status to identify the patients that are unlikely to benefit from gemcitabine based chemotherapies, suggesting that K17 may serve as a predictive biomarker to inform treatment selection. Citation Format: Michael Horowitz, Lyanne Oblein, Jaymie Oentoro, Anika Porwal, Wei Jiang, Natalia Marchenko, Luisa Escobar-Hoyos, Kenneth Shroyer. Keratin 17 and GATA6 expression in pancreatic ductal adenocarcinoma: A subtyping stratification approach [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Pancreatic Cancer; 2023 Sep 27-30; Boston, Massachusetts. Philadelphia (PA): AACR; Cancer Res 2024;84(2 Suppl) nr B066.
oncology
What problem does this paper attempt to address?