The relationship between hypoxia, hypoxia gene signature and survival in patients with breast cancer

Shamis, Suad A. K (2023) The relationship between hypoxia, hypoxia gene signature and survival in patients with breast cancer. PhD thesis, University of Glasgow.

Abstract

Thesis Summary

Breast cancer is the most frequently diagnosed cancer in women in the UK. It is a heterogenous disease with subtypes which behave differently. Although several targeted therapies have been approved for patients with oestrogen receptor (ER) positive and Her-2 positive breast cancer, chemotherapy remains the standard systemic option for triple negative breast cancer (TNBC) patients. New prognostic tools for risk stratification and to guide use of the most aggressive treatments and new targeted therapies are desirable. The role of the tumour microenvironment (TME) in tumour progression is increasingly recognised. Features of the TME such as hypoxia have been reported to have a prognostic role in cancer. A better understanding of this feature may lead to identification of new prognostic and predictive tools and of new therapeutic targets for TNBC.

The work of this thesis is carried out in three cohorts of patients with primary operable breast cancer and mature follow up. Data was available from clinical records regarding patient age, tumour pathology, treatment details and survival. Several hypoxic markers have been investigated and were reported to be overexpressed in breast cancer tissue.

The aim of the current study was to investigate the role of hypoxia inducible factors [HIF1α (1), HIF-1α (2), HIF-2α] and carbonic anhydrases IX (CAIX) in different breast cancer subtypes, to establish biological processes, and key pathways related to cytoplasmic CAIX expression in ER-negative and a node negative subset of ER-negative breast cancer patients, and to identify the mRNA signature associated with CAIX within the tumour and stromal compartments in TNBC.

Immunohistochemistry (IHC) was employed on a tissue microarray (TMA) of patients with breast cancer to assess the expression of HIF-1α (1), HIF-1α (2), HIF-2α and CAIX. Clinical outcomes for the marker’s expression were estimated using Kaplan-Meier analysis and compared between groups with the log-rank test. In a cohort of mixed breast cancer subtypes, the expression of cytoplasmic HIF-1α (1), and HIF-2α were only associated with poor overall survival (OS) in luminal A tumours (P = 0.009), and poor recurrence free survival (RFS) in Her-2 disease (P = 0.038), respectively. However, regardless of cellular localisation, high CAIX expression was associated with poor outcome for the full cohort and in breast cancer subtypes. High cytoplasmic CAIX was associated with decrease RFS in the full cohort (P<0.001), and in luminal B tumours (P = 0.025), disease-free survival (DFS) in the full cohort (P<0.001) in luminal B tumours (P = 0.035), and Her-2 disease (P = 0.016). Also, high cytoplasmic CAIX was associated with poor OS in the full cohort (P<0.001) and in Her-2 disease (P = 0.001). Moreover, membranous CAIX was associated with decrease RFS (P= 0.001), and OS (P = 0.003) in the full cohort. In multivariate analysis, cytoplasmic CAIX was an independent prognostic factor for RFS in the entire cohort (HR = 2.24, 95% CI: 1.19–4.22, P = 0.012), DFS in the full cohort (HR = 1.74, 95% CI: 1.08–2.82, P = 0.023) and in luminal B disease (HR = 3.59, 95% CI: 1.23–10.53, P = 0.020), and OS in Her-2 disease (HR = 4.19, 95% CI: 1.37–12.80, P = 0.012).

Furthermore, in the cohort of ER-positive breast cancer, high cytoplasmic HIF-1α (1) expression was associated with shorter DFS (P = 0.032), and OS (P = 0.002) in the full cohort. In addition, high nuclear HIF-1α (1) expression was associated with decrease DFS in the full cohort (P = 0.009) and in luminal A disease (P = 0.013), and OS in the full cohort (P = 0.002) and in luminal A tumours (P = 0.003). Moreover, high cytoplasmic CAIX expression was corelated with worse RFS and DFS in the full cohort (P = 0.014, 0.008, respectively) and with RFS, DFS, and OS in luminal B disease (P = 0.018, 0.001, 0.003, respectively). On multivariate Cox regression analysis, nuclear HIF-1α (1) was an independent prognostic marker for DFS in the full cohort (HR = 1.85, 95% CI: 1.10–3.11, P = 0.019), and in luminal A disease (HR = 1.98, 95% CI: 1.02–3.83, P = 0.042), and for OS in the full cohort (HR = 1.85, 95% CI: 1.08–3.19, P = 0.026), and in luminal A disease (HR = 2.08, 95% CI: 1.11–3.89, P = 0.022). Moreover, high cytoplasmic CAIX expression was an independent prognostic marker for RFS and DFS in the full cohort (HR = 2.09, 95% CI: 1.17–3.75, P = 0.013; HR = 1.74, 95% CI: 1.08–2.82, P = 0.023), and in luminal B disease (HR = 2.57, 95% CI: 1.29–5.12, P = 0.007; HR = 2.75, 95% CI: 1.66–4.55, P<0.001), respectively.

In TNBC cohort, high cytoplasmic expression of CAIX had lower RFS (P = 0.038). Multivariate analysis showed that cytoplasmic CAIX remained as factor contributing significantly to RFS (HR = 6.59, 95% CI: 1.47–29.58, P = 0.014).

Next, Templated Oligo assay with Sequencing readout (TempO-Seq) (bulk RNAseq) was carried out in ER-negative and a node negative subset of ER-negative breast cancer patients to identify gene signatures that associated with CAIX expression to provide further information on biological processes, and key pathways related to cytoplasmic CAIX expression. Whole transcriptome analysis using TempO-Seq identified 10 significant genes within ER-negative cohort (OR8B2, SERHL2, KRT6A, MMP7, SPINK8, TMEM150C, CEACAM6, MUCL1, PITX2, and GALNT6), and 3 genes in node negative group (PCSK1N, SERHL2 and SPNS2). In node negative patients SPNS2 was of particular interest.

Further, GeoMx Digital Spatial Profiler (DSP) analyses in TNBC cohort was performed to inform if gene signatures were associated with tumour epithelia or TME. 3 upregulated gene expression signatures, CD68, HIF1A, VSIR, and one down-regulated gene, pan-melanocyte, were identified in tumour compartment. In contrast, 9 downregulated genes, CD86, CD3E, MS4A1, BCL2, CCL5, NKG7, PTPRC, CD27 and FAS were identified in the TME in comparison of high and low CAIX expression groups. Among all 4 selected genes, HIF-1α, BCL2, CD68, and CD3, were further validated by IHC at protein level. Univariate KaplanMeier analysis showed high expression of CD68 and HIF-1α was associated with poor prognosis and high expression of BCL2 and CD3 was associated with good prognosis. By performing multivariate analysis for OS, high levels of CD68 cells in tumour nests and in TME were independent prognostic factor for poorer OS (HR = 2.42, 95% CI: 1.05–5.59, P = 0.038; HR = 3.34, 95% CI: 1.28–8.69, P = 0.014), respectively.

In conclusion, this thesis has demonstrated a prognostic role of nuclear HIF-1α (1) and cytoplasmic CAIX in breast cancer. However, their prognostic values were different depending on cellular locations and tumour subtypes. Furthermore, TempO-Seq identified the pathways and genes associated with the CAIX in ER-negative breast cancer. Then, GeoMx DSP technology identified stromal/immune-related genes that were associated with TNBC patients’ survival in comparison of high and low CAIX expression groups that might serve as a potential prognostic biomarker for TNBC.

Overall, the results from this thesis provide new evidence to warrant the further investigation of HIF-1α (1) and CAIX in a large contemporaneous cohort of patients with breast cancer and in particular in patients with TNBC.

• Item Type: Thesis (PhD)
• Qualification Level: Doctoral
• Subjects: Q Science > QH Natural history > QH345 Biochemistry; Q Science > QH Natural history > QH426 Genetics; R Medicine > RC Internal medicine > RC0254 Neoplasms. Tumors. Oncology (including Cancer)
• Colleges/Schools: College of Medical Veterinary and Life Sciences > School of Medicine, Dentistry & Nursing
• Supervisor's Name: McMillan, Professor Donald C. and Edwards, Professor Joanne
• Date of Award: 2023
• Depositing User: Theses Team
• Unique ID: glathesis:2023-83988
• Copyright: Copyright of this thesis is held by the author.
• Date Deposited: 15 Dec 2023 14:21
• Last Modified: 15 Dec 2023 14:25
• Thesis DOI: 10.5525/gla.thesis.83988
• URI: https://theses.gla.ac.uk/id/eprint/83988

Related URLs

• Data DOI
• Article DOI
• Article DOI
• Article DOI
• Article DOI