Ivona Krus1, Veronika Brynychová1, Viktor Hlaváč1,2, Radka Václavíková1, Maria Kováčová3, Renata Koževnikovová4, Katerina Kopečková5, Jannis Tornikidis6, David Vrána7, Jiří Gatěk8, Pavel Souček9,10. 1. Toxicogenomics Unit, National Institute of Public Health, Srobarova 48, 100 42, Prague 10, Czech Republic. 2. Biomedical Center, Faculty of Medicine in Pilsen, Charles University, Pilsen, Czech Republic. 3. Third Faculty of Medicine, Charles University, Prague, Czech Republic. 4. Department of Oncosurgery, MEDICON, Prague, Czech Republic. 5. Department of Oncology, Second Faculty of Medicine, Charles University and Motol University Hospital, Prague, Czech Republic. 6. Department of Surgery, Second Faculty of Medicine, Charles University and Motol University Hospital, Prague, Czech Republic. 7. Comprehensive Cancer Center Novy Jicin, Hospital Novy Jicin, Novy Jicin, Czech Republic. 8. Department of Surgery, EUC Hospital Zlin and Tomas Bata University in Zlin, Zlin, Czech Republic. 9. Toxicogenomics Unit, National Institute of Public Health, Srobarova 48, 100 42, Prague 10, Czech Republic. pavel.soucek@szu.cz. 10. Biomedical Center, Faculty of Medicine in Pilsen, Charles University, Pilsen, Czech Republic. pavel.soucek@szu.cz.
Abstract
INTRODUCTION: Human kinesin 14 (KIF14) is one of the 70 prognostic marker genes (so-called Amsterdam profile) previously identified by the microarray of breast carcinomas, and its high transcript expression in tumor specimens indicates a poor prognosis for patients. We performed a pilot study to explore the prognostic and predictive meaning of KIF14 germline genetic variability in breast cancer patients. METHODS: KIF14 coding sequence, including 5' and 3' untranslated regions and overlaps to introns for identification of splicing sites, was analyzed using next-generation sequencing in the testing set of blood DNA samples from 105 breast cancer patients with clinical follow-up. After rigorous evaluation of major allele frequency, haplotype blocks, in silico predicted functional aspects, expression quantitative trait loci, and clinical associations, eight single nucleotide variants were subsequently validated in the evaluation set of 808 patients. RESULTS: Carriers of minor alleles G (rs17448931) or T (rs3806362) had significantly shorter overall survival than wild type homozygotes (p = 0.010 and p = 0.023, respectively) thus successfully replicating the results of the testing set. Both associations remained significant in the multivariate Cox regression analysis, including molecular subtype and stage as covariates (hazard ratio, HR = 1.7, 95% confidence interval (CI) = 1.1-2.8 for rs17448931 and HR = 1.9, CI 1.2-3.0 for rs3806362). DISCUSSION: In conclusion, our preliminary data suggest that minor alleles in rs17448931 and rs3806362 of KIF14 represent candidate biomarkers of poor prognosis of breast cancer patients. After pending validation in independent populations and eventual functional characterization, these candidates might become useful biomarkers in the clinics.
INTRODUCTION: Human kinesin 14 (KIF14) is one of the 70 prognostic marker genes (so-called Amsterdam profile) previously identified by the microarray of breast carcinomas, and its high transcript expression in tumor specimens indicates a poor prognosis for patients. We performed a pilot study to explore the prognostic and predictive meaning of KIF14 germline genetic variability in breast cancer patients. METHODS: KIF14 coding sequence, including 5' and 3' untranslated regions and overlaps to introns for identification of splicing sites, was analyzed using next-generation sequencing in the testing set of blood DNA samples from 105 breast cancer patients with clinical follow-up. After rigorous evaluation of major allele frequency, haplotype blocks, in silico predicted functional aspects, expression quantitative trait loci, and clinical associations, eight single nucleotide variants were subsequently validated in the evaluation set of 808 patients. RESULTS: Carriers of minor alleles G (rs17448931) or T (rs3806362) had significantly shorter overall survival than wild type homozygotes (p = 0.010 and p = 0.023, respectively) thus successfully replicating the results of the testing set. Both associations remained significant in the multivariate Cox regression analysis, including molecular subtype and stage as covariates (hazard ratio, HR = 1.7, 95% confidence interval (CI) = 1.1-2.8 for rs17448931 and HR = 1.9, CI 1.2-3.0 for rs3806362). DISCUSSION: In conclusion, our preliminary data suggest that minor alleles in rs17448931 and rs3806362 of KIF14 represent candidate biomarkers of poor prognosis of breast cancer patients. After pending validation in independent populations and eventual functional characterization, these candidates might become useful biomarkers in the clinics.