Molecular profiling of platinum resistant ovarian cancer Jozien Helleman 1 , Maurice P.H.M. Jansen 1 , Paul N. Span 2 , Iris L. van Staveren 1 , Leon F.A.G. Massuger 3 , Marion E. Meijer-van Gelder 1 , Fred C.G.J. Sweep 2 , Patricia C. Ewing 4 , Maria E.L. van der Burg 1 , Gerrit Stoter 1 , Kees Nooter 1 and Els M.J.J. Berns 1 * 1 Department of Medical Oncology, Erasmus MC/Daniel den Hoed Cancer Centre, Rotterdam, The Netherlands 2 Department of Chemical Endocrinology, Radboud University Nijmegen Medical Centre, Nijmegen, The Netherlands 3 Department of Obstetrics and Gynecology, Radboud University Nijmegen Medical Centre, Nijmegen, The Netherlands 4 Department of Pathology, Erasmus MC/Daniel den Hoed Cancer Centre, Rotterdam, The Netherlands The aim of this study is to discover a gene set that can predict resistance to platinum-based chemotherapy in ovarian cancer. The study was performed on 96 primary ovarian adenocarcinoma specimens from 2 hospitals all treated with platinum-based che- motherapy. In our search for genes, 24 specimens of the discovery set (5 nonresponders and 19 responders) were profiled in duplicate with 18K cDNA microarrays. Confirmation was done using quan- titative RT-PCR on 72 independent specimens (9 nonresponders and 63 responders). Sixty-nine genes were differentially expressed between the nonresponders (n 5 5) and the responders (n 5 19) in the discovery phase. An algorithm was constructed to identify pre- dictive genes in this discovery set. This resulted in 9 genes (FN1, TOP2A, LBR, ASS, COL3A1, STK6, SGPP1, ITGAE, PCNA), which were confirmed with qRT-PCR. This gene set predicted platinum resistance in an independent validation set of 72 tumours with a sensitivity of 89% (95% CI: 0.68–1.09) and a specificity of 59% (95% CI: 0.47–0.71)(OR 5 0.09, p 5 0.026). Multivariable analysis including patient and tumour characteristics demon- strated that this set of 9 genes is independent for the prediction of resistance (p < 0.01). The findings of this study are the discovery of a gene signature that classifies the tumours, according to their response, and a 9-gene set that determines resistance in an inde- pendent validation set that outperforms patient and tumour char- acteristics. A larger independent multicentre study should further confirm whether this 9-gene set can identify the patients who will not respond to platinum-based chemotherapy and could benefit from other therapies. ' 2005 Wiley-Liss, Inc. Key words: ovarian cancer; cDNA microarray; cisplatin; carbopla- tin; response prediction Ovarian cancer is the leading cause of death from gynecological cancers in the western world 1 and is the fifth most frequent cause of cancer death in women, with 1,100 new cases each year in the Netherlands 2 and 192,000 cases worldwide. 3 Ninety percent of malignant ovarian tumours are ovarian adenocarcinomas. 4 The treatment of ovarian adenocarcinoma has improved over the last 20 years due to improved debulking surgery and chemotherapy, 5 especially since the introduction of platinum-based drugs and, more recently, the addition of taxanes. 4,5 Despite these treatment improvements, 20–30% of patients never have a clinical remission and the majority of women will eventually relapse with generally incurable disease. 6 Although optimal debulking surgery 7,8 and early stage of dis- ease 8 are associated with a better (recurrence free) survival, it is impossible to predict which patient will progress or recur during or after chemotherapy. This prediction is essential since patients that are resistant might benefit from a different combinational treatment. Moreover, a better understanding of the platinum resist- ance mechanism is needed for response prediction as well as for development of drugs that could circumvent resistance mecha- nisms. Microarray technology has given us the ability to determine the expression of thousands of genes in a single experiment, and has provided an opportunity to classify different subtypes of cancer based on characteristic expression patterns. Several gene expres- sion profiling studies on ovarian cancer specimens identified genes that can be useful as molecular markers for a better diagnosis (e.g., HE4, Mucin1 and Meis1) 9–13 or gene-signatures that can dis- tinguish between different histological subtypes. 2,14,15 Moreover, the microarray technology proved to be useful in the discovery of genes associated with platinum resistance in a panel of 14 ovarian cancer cell lines 16 and genes associated with the development of platinum or paclitaxel resistance in lung cancer cell lines 17 and ovarian cancer cell lines, 18 respectively. Recently, a 14-gene sig- nature was discovered that predicts early relapse in ovarian cancer after platinum-paclitaxel chemotherapy, but not the response. 19 The aim of our study is to find a gene set that can predict the resistance to platinum-based chemotherapy and that might lead to novel targets for therapeutic intervention. Primary tumour speci- mens of 96 ovarian adenocarcinoma patients were included in this study. All patients were treated with platinum-based chemother- apy and the main clinical endpoint of this study is the resistance to platinum-based treatment. Material and methods Patients The study design was approved by the Medical Ethical Com- mittee of the Erasmus MC Rotterdam, the Netherlands (MEC 02.949). Ninety-six primary ovarian cancer specimens (i.e. before chemotherapy) were used, originating from 2 hospitals, Erasmus MC in Rotterdam (n 5 50) and the Radboud University Nijmegen Medical Centre (Radboud UNMC) (n 5 46). The patient and tumour characteristics are listed in Table I. Eighty-four percent of the patients received cisplatin or carboplatin in combination with endoxan (including all patients of the discovery set) and 16% of the patients were treated with other platinum-based chemotherapy. Overall, 82 patients responded to chemotherapy, whereas 14 did not, which is comparable with the response rate of 80% seen in the clinic. The clinical response was assessed according to the standard WHO response criteria. Complete response was defined as the disappearance of all clinically measurable tumour lesions. Partial response was defined as a 50% or more decrease of all lesions. Stable disease was either a decrease in size of less than 50% or an increase in size of less than 25% of one or more meas- ured tumour lesions. Progressive disease was either a 25% or more increase in size of one or more clinically measured lesions or the appearance of new disease. In detail, 43 patients had a complete response, 15 partial (these tumours may contain a subpopulation of resistant cells) and 1 stable disease, and 18 patients had no Presented in part at the CNIO 2004 Symposium on The Molecular Taxonomy of Cancer, Madrid (poster presentation) and at the 96th AACR Annual Meeting, Anaheim, CA (oral presentation). Grant sponsor: Erasmus MC Revolving Fund; Grant sponsor: Dutch Cancer Society, Amsterdam, the Netherlands; Grant number: DDHK 2364. *Correspondence to: Erasmus MC, Department of Medical Oncology, Josephine Nefkens Institute, Room Be424, P.O. Box 1738, 3000 DR, The Netherlands. Fax: 131-10-408-8377. E-mail: p.berns@erasmusmc.nl Received 8 July 2005; Accepted after revision 29 August 2005 DOI 10.1002/ijc.21599 Published online 14 November 2005 in Wiley InterScience (www. interscience.wiley.com). Int. J. Cancer: 118, 1963–1971 (2006) ' 2005 Wiley-Liss, Inc. Publication of the International Union Against Cancer