Integrated genome-wide DNA copy number and expression analysis identifies distinct mechanisms of primary chemoresistance in ovarian carcinomas
October 2009
Commentary by Dr Izhak Haviv
The article:
Etemadmoghadam D, DeFazio A, Beroukhim R, et al. Integrated genome-wide DNA copy number and expression analysis identifies distinct mechanisms of primary chemoresistance in ovarian carcinomas. Clin Cancer Res 2009;15(4):1417-27.
The reviewer:
Dr Haviv is Head of System Integrations, Blood & DNA Profiling Facility, Baker Institute Melbourne. Dr Haviv also works in Department of Biochemistry, School of Medicine, University of Melbourne and Peter MacCallum Cancer Centre.
Abbreviations
Cyclin E (CCNE1)
Summary
Study Design
A significant number of women with serous ovarian cancer exhibit resistance to platinum-based treatment. This study analysed somatic DNA copy number variation and gene expression data to identify key mechanisms associated with primary resistance in advanced-stage serous cancers.
High-resolution oligonucleotide copy number analysis was performed on 118 ovarian tumours from patients in Australia, Norway and Japan. A selected subset of 85 advanced-stage serous tumours from 52 patients responsive to platinum-based treatment and 33 resistant patients was used to relate copy number variation to primary response to treatment. This approach was complemented by quantitative-PCR copy number analysis of 12 candidate genes previously associated in other studies with clinical course in ovarian cancer. Gene expression profiling was used to investigate the molecular changes in tumours that differed in amplification of a candidate region that was identified, and response to treatment.
Findings
Copy number analysis indicated that amplification of 19q12, containing cyclin E (CCNE1), and 20q11.22-q13.12, mapping adjacent to the steroid receptor coactivator NCOA3, was significantly associated with primary treatment resistance. The association between CCNE1 amplification and clinical outcome was confirmed, whereas in this study other genes previously associated with copy number variation and clinical outcome in ovarian cancer, were not associated with primary treatment resistance.
Gene expression profiling and immunohistochemistry indicated chemoresistant tumours with high CCNE1 copy number and protein expression were associated with increased cellular proliferation, however cellular proliferation alone may not account for the response to chemotherapy. Patients with a poor clinical outcome, without CCNE1 amplification, over-expressed genes involved in extracellular matrix deposition.
Conclusion
Two distinct mechanisms of primary treatment failure in serous ovarian cancer were identified: mechanisms involving CCNE1 amplification and enhanced extracellular matrix deposition. The validation of CCNE1 copy number as a dominant marker of patient outcome in ovarian cancer, suggests the potential for cyclin-related targeted treatments for patients with CCNE1 amplification.
Commentary
What does this article add to existing clinical evidence in this area?
The statement made in the manuscript claiming the study is “the most robust analysis of primary response in ovarian cancer to date” is well founded, both in terms of confidence as well as how clean the comparison was from confounding factors.
Studies that seek associations between genes and clinical features on a genomic scale may suffer from several design flaws that hinder the interpretation of the data.1 Firstly, the vast heterogeneity of cancer genomes introduces the risk of identifying false associations, based on the samples chosen. Secondly, as the samples compared are different in many ways, often the identified genes actually represent disparity in well known modifiers of the clinical dichotomy being investigated. In this study, smaller sets of samples were compared which exhibit parity for all recognizable clinical features of the samples, other than the study focus (in this case, response variables of overall and disease free survival). Parity is also required in the level of aetiological homogeneity. In the case of ovarian cancer, histopathological subtypes of ovarian cancer should each be considered as independent diseases. Achieving the sample numbers that would allow parity and homogeneity is not trivial.
In addition to satisfying all critical design requirements, the high quality of this study is reflected by the extensive benchmarking of their discovery in light of existing evidence in the literature. Furthermore, with the benefit of the sample size in this study, past discoveries could be assessed with the current samples.
How adequate was the methodology used in addressing the aims of this study?
While other studies in the field of chemotherapy response may exceed this study for either sample numbers or genomic resolution, this study has the largest combined power of resolution and sample number. In short, this manuscript is a gold standard model for genomic association studies, and offers novel analysis tools that are superior to existing tools, specifically for the analysis of copy number variations across large number of samples.
To identify drug-resistance associated lesions requires enormous numbers of samples to be analysed via international collaborations, such as that in this study, since cancer genomes are so extensively aberrant. The number of genomic aberrations observed in cancer samples far exceeds the number of phenotypic milestones cancer exhibits. This is partially due to massive numbers of non-functional mutations associated with the defective genomic integrity mechanisms in cancer cells and is also contributed to by extensive redundancy in the genomic hardwire of molecular control mechanisms and signal transduction. This means that lesions may occur in multiple foci and that additional genes, other than cyclin E, may contribute to the chemo-resistance, however insufficient evidence has been obtained to vindicate them. Beyond the identification of genomic regions, the mechanisms still need elucidation for the data to be useful for drug development.
By combining expression data with its potential underlying mechanism of chromosomal aberration, the authors were able to pinpoint the genes that are most likely to provide the selective advantage to the cancer cell, as reflected by the increased frequency of the aberration in the population. In summary, the study is more than adequate, and serves as a platform for further investigation of ovarian cancer resistance to current treatment in many ways.
What are the implications of this study for clinical practice in Australia?
Expanding our understanding of how cancer resists current treatment using robust tools such as whole genome profiling, bears the potential for better patient stratification (using prognostic markers for response) as well as novel therapeutic avenues. It is hard to envisage or justify a decision to avoid carboplatin-taxol treatment of a patient, merely because her case exhibits cyclin E amplification. However, further mechanistic insight to the biology of resistance, is key to the development of more effective modalities. Indeed, the emerging paradigm of recent targeted drugs, such as trastuzumab, turns defined drug-resistance associated lesions, into the cancer’s Achilles heel. Due to extensive feedback networks hardwired into our genome programs, any given cancer lesion simultaneously enhances cell proliferation as well as apoptosis. This paradigm of “oncogene addiction” of cancer cells2 was originally conceived over the feedback loops involving cyclin E.
For this major achievement to fulfil its promise to improve existing clinical practices, further pursuit of cyclin E targeting drugs is needed. While this vision of personalised medicine of ovarian cancer is not yet a reality, this report by Etemadmoghadam et al certainly provides a key navigation sign for such pursuit.
References
- Tinker AV, Boussioutas A, Bowtell D. The challenges of gene expression microarrays for the study of human cancer. Cancer Cell 2006;9(5):333-9.
- Weinstein IB, Begemann M, Zhou P, et al. Disorders in cell circuitry associated with multistage carcinogenesis: exploitable targets for cancer prevention and therapy Clinical Cancer Research 1997;3:2696-702.
Editor: Dr Anne Nelson, Evidence Review and Research Leader, National Breast and Ovarian Cancer Centre.
Clinical Update - Ovarian Cancer Editorial Committee: Prof Michael Friedlander – Medical Oncologist, Prof Neville Hacker – Gynaecological Oncologist, Ms Kim Hobbs – Social Worker, Dr Gillian Mitchell – Medical Oncologist, Dr Deborah Neesham – Gynaecological Oncologist, Ms Georgie Richter – Gynaecological Nurse.
Disclaimer
Clinical Update - Ovarian Cancer is produced by the National Breast and Ovarian Cancer Centre (NBOCC) and is intended to provide health professionals with timely expert commentary on new research in ovarian cancer. Commentaries included in Clinical Update - Ovarian Cancer do not replace recommendations included in NBOCC clinical practice guidelines.
Information contained in Clinical Update - Ovarian Cancer is not intended to be used as substitute for an independent health professional's advice. The NBOCC does not accept any liability for any injury, loss or damage incurred by use of or reliance on the information contained in Clinical Update - Ovarian Cancer. The NBOCC develops material based on the best available evidence however cannot guarantee and assumes no legal liability or responsibility for the currency or completeness of the information.