Advertisement

Predictive biomarkers of platinum and taxane resistance using the transcriptomic data of 1816 ovarian cancer patients

  • János Tibor Fekete
    Affiliations
    Semmelweis University, Department of Bioinformatics, H-1094 Budapest, Hungary

    Research Center for Natural Sciences, Momentum Cancer Biomarker Research Group, Institute of Enzymology, Magyar tudósok körútja 2, H-1117 Budapest, Hungary

    Semmelweis University, 2nd Department of Pediatrics, H-1094 Budapest, Hungary
    Search for articles by this author
  • Ágnes Ősz
    Affiliations
    Semmelweis University, Department of Bioinformatics, H-1094 Budapest, Hungary

    Research Center for Natural Sciences, Momentum Cancer Biomarker Research Group, Institute of Enzymology, Magyar tudósok körútja 2, H-1117 Budapest, Hungary

    Semmelweis University, 2nd Department of Pediatrics, H-1094 Budapest, Hungary
    Search for articles by this author
  • Imre Pete
    Affiliations
    National Institute of Oncology, H-1122 Budapest, Hungary
    Search for articles by this author
  • Gyula Richárd Nagy
    Affiliations
    Research Center for Natural Sciences, Momentum Cancer Biomarker Research Group, Institute of Enzymology, Magyar tudósok körútja 2, H-1117 Budapest, Hungary

    Semmelweis University, Department of Obstetrics and Gynecology, H-1085 Budapest, Hungary
    Search for articles by this author
  • Ildikó Vereczkey
    Affiliations
    National Institute of Oncology, H-1122 Budapest, Hungary
    Search for articles by this author
  • Balázs Győrffy
    Correspondence
    Corresponding author at: Semmelweis University, Department of Bioinformatics, Tűzoltó u. 7-9, 1094 Budapest, Hungary.
    Affiliations
    Semmelweis University, Department of Bioinformatics, H-1094 Budapest, Hungary

    Research Center for Natural Sciences, Momentum Cancer Biomarker Research Group, Institute of Enzymology, Magyar tudósok körútja 2, H-1117 Budapest, Hungary

    Semmelweis University, 2nd Department of Pediatrics, H-1094 Budapest, Hungary
    Search for articles by this author
Open AccessPublished:January 20, 2020DOI:https://doi.org/10.1016/j.ygyno.2020.01.006

      Highlights

      • Biomarkers capable to predict response to platinum and taxane combination treatment in ovarian tumors were identified.
      • Proteins including NCOR2, TFE3, AKIP1 and AKIRIN2 were among the most significant genes validated in an independent cohort.
      • The integrated database with available treatment and response data can be mined to validate new biomarker candidates.

      Abstract

      Objective

      The first-line chemotherapy for ovarian cancer is based on a combination of platinum and taxane. To date, no reliable predictive biomarker has been recognized that is capable of identifying patients with pre-existing resistance to these agents. Here, we have established an integrated database and identified the most significant biomarker candidates for chemotherapy resistance in serous ovarian cancer.

      Methods

      Gene arrays were collected from the GEO and TCGA repositories. Treatment response was defined based on pathological response or duration of relapse-free survival. The responder and nonresponder cohorts were compared using the Mann-Whitney and receiver operating characteristic tests. An independent validation set was established to investigate the correlation between chemotherapy response for the top 8 genes. Statistical significance was set at p < 0.05.

      Results

      The entire database included 1816 tumor samples from 12 independent datasets. From analyzing all the genes for platinum + taxane response, we identified the eight strongest genes correlated to chemotherapy resistance: AKIP1 (p = 1.60E-08, AUC = 0.728), MARVELD1 (p = 2.70E-07, AUC = 0.712), AKIRIN2 (p = 2.60E-07, AUC = 0.704), CFL1 (p = 8.10E-08, AUC = 0.694), SERBP1 (p = 8.10E-07, AUC = 0.684), PDXK (p = 1.30E-04, AUC = 0.634), TFE3 (p = 7.90E-05, AUC = 0.631) and NCOR2 (p = 1.90E-03, AUC = 0.611). Of these, the independent validation confirmed TFE3 (p = 0.012, AUC = 0.718), NCOR2 (p = 0.048, AUC = 0.671), PDXK (p = 0.019, AUC = 0.702), AKIP1 (p = 0.002, AUC = 0.773), MARVELD1 (p = 0.044, AUC = 0.675) and AKIRIN2 (p = 0.042, AUC = 0.676). An online interface was set up to enable future validation and ranking of new biomarker candidates in an automated manner (www.rocplot.org/ovar).

      Conclusions

      We compiled a large integrated database with available treatment and response information and used this to uncover new biomarkers of chemotherapy response in serous ovarian cancer.

      Keywords

      1. Introduction

      Ovarian cancer is the second most common cause of death related to gynecologic malignancies in women. In 2018, the number of new cases worldwide was estimated to be 295,414 and the number of deaths was estimated to be 184,799 [
      • Bray F.
      • Ferlay J.
      • Soerjomataram I.
      • Siegel R.L.
      • Torre L.A.
      • Jemal A.
      Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries.
      ]. The incidence and mortality rates of the disease vary by region and have shown an increasing trend in developed countries in the last decade [
      • Menyhart O.
      • Fekete J.T.
      • Gyorffy B.
      Demographic shift disproportionately increases cancer burden in an aging nation: current and expected incidence and mortality in Hungary up to 2030.
      ]. Almost 90% of ovarian malignancies are of epithelial origin. From a histopathological point of view, epithelial ovarian cancer (EOC) can be divided into serous, endometroid, mucinous and clear-cell histology subtypes. Of these, high-grade serous carcinoma is the most frequently diagnosed type [
      • Matulonis U.A.
      • Sood A.K.
      • Fallowfield L.
      • Howitt B.E.
      • Sehouli J.
      • Karlan B.Y.
      Ovarian cancer.
      ].
      Early diagnosis of ovarian cancer is difficult because most patients are asymptomatic in the early stages, and therefore, many tumors are detected only in the advanced stages. The recommendations of The National Comprehensive Cancer Network (NCCN) for advanced ovarian cancer treatment comprise surgery followed by systemic chemotherapy using cisplatin/paclitaxel (www.nccn.org). Although more than 80% of the patients initially respond to first-line treatment, most will have a recurrence within two years that progresses to advanced disease [
      • Ozols R.F.
      Systemic therapy for ovarian cancer: current status and new treatments.
      ].
      The most established biomarker for ovarian cancer is serum CA125, which is the gold standard for disease monitoring. A correlation between CA125 and chemotherapy response was proposed earlier [
      • Crawford S.M.
      • Peace J.
      Does the nadir CA125 concentration predict a long-term outcome after chemotherapy for carcinoma of the ovary?.
      ], but it has not yet reached clinical application. Conversely, there is currently no validated predictive biomarker for ovarian cancer, although there is an urgent need to identify patients who are unlikely to benefit from platinum and taxane combination therapy.
      To deliver personalized treatment decisions, a clear understanding of the exact mechanisms of drug resistance is required; resistance can be caused by multiple independent mechanisms, some of which have been extensively studied in patient samples and cell culture model systems. We and others have shown previously that resistance mechanisms against platinum-based agents include decreased drug accumulation [
      • Samimi G.
      • Varki N.M.
      • Wilczynski S.
      • Safaei R.
      • Alberts D.S.
      • Howell S.B.
      Increase in expression of the copper transporter ATP7A during platinum drug-based treatment is associated with poor survival in ovarian cancer patients.
      ], drug inactivation by glutathione and glutathione S-transferases [
      • Pljesa-Ercegovac M.
      • Savic-Radojevic A.
      • Matic M.
      • Coric V.
      • Djukic T.
      • Radic T.
      • Simic T.
      Glutathione Transferases: potential targets to overcome chemoresistance in solid tumors.
      ], increased autophagy [
      • Wang J.
      • Wu G.S.
      Role of autophagy in cisplatin resistance in ovarian cancer cells.
      ], and increased levels of DNA repair [
      • Penzvalto Z.
      • Surowiak P.
      • Gyorffy B.
      Biomarkers for systemic therapy in ovarian cancer.
      ]. Taxane resistance can result from overexpressed efflux pump genes [
      • Nanayakkara A.K.
      • Follit C.A.
      • Chen G.
      • Williams N.S.
      • Vogel P.D.
      • Wise J.G.
      Targeted inhibitors of P-glycoprotein increase chemotherapeutic-induced mortality of multidrug resistant tumor cells.
      ], modulated microtubule dynamics [
      • Orr G.A.
      • Verdier-Pinard P.
      • McDaid H.
      • Horwitz S.B.
      Mechanisms of Taxol resistance related to microtubules.
      ], altered expression of β tubulin isotypes [
      • Orr G.A.
      • Verdier-Pinard P.
      • McDaid H.
      • Horwitz S.B.
      Mechanisms of Taxol resistance related to microtubules.
      ] and enhanced epithelial-to-mesenchymal transition [
      • Kampan N.C.
      • Madondo M.T.
      • McNally O.M.
      • Quinn M.
      • Plebanski M.
      Paclitaxel and its evolving role in the management of ovarian cancer.
      ]. Multiple studies have described gene expression alterations that are associated with drug resistance in ovarian cancer. Higher expression of IGF2BP [
      • Hsu K.F.
      • Shen M.R.
      • Huang Y.F.
      • Cheng Y.M.
      • Lin S.H.
      • Chow N.H.
      • Cheng S.W.
      • Chou C.Y.
      • Ho C.L.
      Overexpression of the RNA-binding proteins Lin28B and IGF2BP3 (IMP3) is associated with chemoresistance and poor disease outcome in ovarian cancer.
      ], LIN28B [
      • Hsu K.F.
      • Shen M.R.
      • Huang Y.F.
      • Cheng Y.M.
      • Lin S.H.
      • Chow N.H.
      • Cheng S.W.
      • Chou C.Y.
      • Ho C.L.
      Overexpression of the RNA-binding proteins Lin28B and IGF2BP3 (IMP3) is associated with chemoresistance and poor disease outcome in ovarian cancer.
      ] and MSLN [
      • Cheng W.F.
      • Huang C.Y.
      • Chang M.C.
      • Hu Y.H.
      • Chiang Y.C.
      • Chen Y.L.
      • Hsieh C.Y.
      • Chen C.A.
      High mesothelin correlates with chemoresistance and poor survival in epithelial ovarian carcinoma.
      ] were reported in paclitaxel- and platinum-treated nonresponder tissues. Upregulated CHI3L1 inhibited paclitaxel-induced apoptosis in nonresponder cell lines [
      • Januchowski R.
      • Swierczewska M.
      • Sterzynska K.
      • Wojtowicz K.
      • Nowicki M.
      • Zabel M.
      Increased expression of several collagen genes is associated with drug resistance in ovarian cancer cell lines.
      ]. The higher expression of FOXM1 increased cell cycle progression in platinum-treated drug-resistant tissue samples [
      • Chen C.
      • Chang Y.C.
      • Lan M.S.
      • Breslin M.
      Leptin stimulates ovarian cancer cell growth and inhibits apoptosis by increasing cyclin D1 and Mcl-1 expression via the activation of the MEK/ERK1/2 and PI3K/Akt signaling pathways.
      ,
      • Lane D.
      • Matte I.
      • Garde-Granger P.
      • Laplante C.
      • Carignan A.
      • Rancourt C.
      • Piche A.
      Inflammation-regulating factors in ascites as predictive biomarkers of drug resistance and progression-free survival in serous epithelial ovarian cancers.
      ]. In cisplatin-resistant cell lines, upregulated CSF-1R [
      • Yu R.
      • Jin H.
      • Jin C.
      • Huang X.
      • Lin J.
      • Teng Y.
      Inhibition of the CSF-1 receptor sensitizes ovarian cancer cells to cisplatin.
      ] and downregulated OXCT1 facilitated the inhibition of apoptosis [
      • Yang S.D.
      • Ahn S.H.
      • Kim J.I.
      3-Oxoacid CoA transferase 1 as a therapeutic target gene for cisplatin-resistant ovarian cancer.
      ,
      • Sanchez-Alvarez R.
      • Martinez-Outschoorn U.E.
      • Lin Z.
      • Lamb R.
      • Hulit J.
      • Howell A.
      • Sotgia F.
      • Rubin E.
      • Lisanti M.P.
      Ethanol exposure induces the cancer-associated fibroblast phenotype and lethal tumor metabolism: implications for breast cancer prevention.
      ].
      In the present study, our aim was to establish a framework to uncover and validate gene expression-based predictive biomarkers of therapy resistance by mining large, publicly available transcriptomic datasets of ovarian cancer patients with known treatment protocols and available clinical follow-up data. Furthermore, we also performed an independent sample collection of ovarian cancer specimens and performed RT-PCR using RNA from these tumor samples to validate the best performing biomarker candidates for predicting platinum and taxane resistance.

      2. Methods

      2.1 Database construction

      We searched GEO (http://www.pubmed.com/geo) and TCGA (http://cancergenome.nih.gov) repositories to identify datasets suitable for the analysis. In this search, the keywords “ovarian”, “cancer”, “treatment”, “response”, and “survival” were used. Only publications with available raw microarray gene expression data, clinical treatment, response or survival information, and at least 20 patients were included. Only three closely related microarray platforms, GPL96 (Affymetrix HG-U133A), GPL570 (Affymetrix HG-U133 Plus 2.0), and GPL571/GPL3921 (Affymetrix HG-U133A 2.0), were considered.

      2.2 Preprocessing

      First, the raw .CEL files were MAS5 normalized in the R statistical environment (www.r-project.org) using the Affy Bioconductor library [
      • Gautier L.
      • Cope L.
      • Bolstad B.M.
      • Irizarry R.A.
      affy—analysis of Affymetrix GeneChip data at the probe level.
      ]. This was followed by a second scaling normalization to set the average expression of the 22,277 identical probe sets in each chip to 1000 [
      • Sims A.H.
      • Smethurst G.J.
      • Hey Y.
      • Okoniewski M.J.
      • Pepper S.D.
      • Howell A.
      • Miller C.J.
      • Clarke R.B.
      The removal of multiplicative, systematic bias allows integration of breast cancer gene expression datasets - improving meta-analysis and prediction of prognosis.
      ]. Normalized gene expression and clinical data were integrated into a PostgreSQL relational database.

      2.3 Statistical computations

      The tumor samples were divided into responder and nonresponder cohorts based on their clinical characteristics. For cases with available pathological response (PR), we classified the patients as published by the authors (PR dataset). If the pathological response was not available, the classification was based on the duration of the progression-free survival. Those with a relapse-free survival shorter than six months were compared to those without a relapse before six months. Patients censored before six months were excluded from the analysis.
      The two cohorts were compared using the Mann-Whitney test and the receiver operating characteristic test in the R statistical environment (www.r-project.org) using Bioconductor libraries (www.bioconductor.org). The cutoff for p values was set at p < 0.05. False Discovery Rate (FDR) was calculated using the q value package (http://github.com/jdstorey/qvalue), and only results with a FDR < 5% were accepted as significant.

      2.4 Clinical sample collection

      In total, 81 fresh frozen ovarian tissue samples were collected during surgery from patients with ovarian cancer at the National Institute of Cancer (OOI) Budapest, Hungary (OOI set). Samples were stored in RNA later (Thermo Fisher Scientific, USA) at −80 °C until RNA isolation. An institutional ethics committee (Országos Onkológiai Intézet, Intézeti Kutatásetikai Bizottság - OOI IKEB) approved the study with the reference number OOI-Ált-9444-1/2013/59. Anonymized clinical data were obtained from medical and pathological records.

      2.5 RNA isolation and cDNA synthesis

      Total RNA was isolated using the AllPrep DNA/RNA Kit (Qiagen, Germany) following the manufacturer's protocol. The quality of RNA was assessed by UV spectrophotometry (NanoDrop, Thermo Fisher Scientific, USA). For quantitative PCR analysis, 1 μg of total RNA was reverse transcribed in a final volume of 20 μl using the Maxima First Strand cDNA Synthesis Kit for RT-qPCR, and dsDNase was used to remove any potential DNA contamination (Thermo Fisher Scientific, USA).

      2.6 Quantitative PCR analysis

      Quantitative PCR was performed in a CFX384 real-time PCR instrument (Bio-Rad Laboratories, USA) using the SensiFAST SYBR No-ROX Kit (Bioline Reagents, UK). Primers were designed for the same exons targeted by the microarray probes of each selected gene. GAPDH and ACTB were employed as endogenous controls for normalization. The reactions were performed in 10 μl containing 1 μl of cDNA, diluted 25-fold, and 250 nM of each primer. After an initial denaturation step for 2 min at 95 °C, 36 cycles with three steps were performed: 95 °C for 10 s, 62 °C for 10 s and 72 °C for 20 s. Each sample was measured in triplicate, and the threshold cycle (Ct) was determined for all genes using Bio-Rad CFX Maestro software (Bio-Rad Laboratories, USA). Relative gene expression values were analyzed using the ΔCt method.

      2.7 Validation

      For the validation, we selected a set of the best performing genes from the taxane- and platinum-treated samples. Similar to the discovery set, the Mann-Whitney test and ROC analyses were performed to compare the expression of each investigated gene in the responder and nonresponder cohorts. Statistical significance was set at p < 0.05.

      3. Results

      3.1 Database

      Overall, 10,283 patients in 134 GEO and TCGA datasets met our search criteria (Fig. 1A ). After eliminating samples with insufficient clinical data, we maintained 1816 ovarian cancer samples. Of these, pathologic response was available for 1022 patients (PR dataset) and relapse-free survival at 6 months was obtainable for 1347 patients (RFS dataset). We selected the RFS dataset for further analysis because this included the larger patient cohort. Aggregate clinical characteristics of the cohorts are presented in Table 1 and Fig. 1B.
      Fig. 1
      Fig. 1Overview of the ovarian cancer databases included in the study. (A) Flowchart for the setup of the discovery datasets. (B) Clinical characteristics of each included dataset. OOI = National Institute of Oncology.
      Table 1Summary of the clinical characteristics of the datasets included in the analysis. PCR: pathological response; RFS: relapse-free survival at 6 months.
      DatasetPlatformSample sizeHistologyGradeStageDebulk
      PCRRFS 6Serous/endometrioid/clear cell1/2/3/4I/II/III/IVProportion of optimal debulking
      GSE14764GPL96808068/7/−3/23/54/−8/1/69/293%
      GSE15622GPL571353531/−/−−/7/28/−−/−/26/9
      GSE26712GPL9629179/8/67/33/67/−20/11/59/1744%
      GSE30161GPL570555345/−/−2/19/30/−−/−/50/543%
      GSE32062GPL5701010/−/−−/3/7/−−/−/6/440%
      GSE51373GPL570282828/−/−−/−/−/−−/5/19/3
      GSE63885GPL5707570/1/−−/9/48/18−/2/63/1020%
      GSE65986GPL5705116/14/22−/−/−/−23/4/11/9
      GSE9891GPL570230272252/19/119/93/156/−21/17/208/2270%
      GSE23603GPL962828/−/−2/8/18/−−/−/−/−48%
      GSE3149GPL96116−/−/−4/55/53/1−/1/95/1954%
      TCGAGPL3921450452452/−/−5/60/378/111/24/349/6673%
      Total102213471079/49/2942/310/839/2083/65/955/16659%

      3.2 Treatment cohorts

      Most of the patients in the RFS cohort received platinum therapy (n = 1209), and more than 50% of the patients were administered taxane (n = 888). Other groups of patients received gemcitabine (n = 126), topotecan (n = 118), paclitaxel (n = 869), docetaxel (n = 97), or Avastin (n = 50). Treatment cohorts in the RFS and PR datasets are summarized in Table 2.
      Table 2Overview of treatments administered to patients included in the discovery datasets.
      TreatmentRelapse-free survival at 6 monthsPathological response
      Platinum1209961
      Taxane888851
      Platinum & taxane871834
      Docetaxel9799
      Paclitaxel869828
      Gemcitabine126121
      Topotecan118108
      Avastin5047
      Total13471022

      3.3 Predictive biomarker candidates

      We filtered for patients who received platinum and taxane combination therapy. Furthermore, as the number of available samples from endometrioid (n = 24) and clear cell (n = 8) histology subtypes were limited, we retained only samples with serous histology and then performed the Mann-Whitney test and ROC analysis for all genes in the RFS datasets. The top eight genes with the highest area under the curve (AUC) values were selected for further validation. The selected genes were AKIP1 (p = 1.60E-08, AUC = 0.728), MARVELD1 (p = 2.70E-07, AUC = 0.712), AKIRIN2 (p = 2.60E-07, AUC = 0.704), CFL1 (p = 8.10E-08, AUC = 0.694), SERBP1 (p = 8.10E-07, AUC = 0.684), PDXK (p = 1.30E-04, AUC = 0.634), TFE3 (p = 7.90E-05, AUC = 0.631) and NCOR2 (p = 1.90E-03, AUC = 0.611). Detailed results are presented in Table 3 and Fig. 2.
      Table 3Top 8 gene expression-based biomarker candidates of platinum + taxane combined chemotherapy response in the RFS discovery dataset.
      Affymetrix IDGene symbolnAUCAUC p-valueMann-Whitney test p-valueFalse discovery rateMedian fold change
      242515_x_atAKIP13360.7281.60E-083.40E-071.68E-051.37
      223095_atMARVELD13360.7122.70E-072.20E-061.62E-041.64
      223144_s_atAKIRIN23360.7042.60E-074.90E-061.59E-041.77
      200021_atCFL18180.6948.10E-088.10E-087.21E-051.47
      227369_atSERBP13360.6848.10E-074.00E-053.22E-041.48
      218019_s_atPDXK8180.6341.30E-042.20E-048.16E-031.46
      212457_atTFE38180.6317.90E-053.00E-045.82E-031.24
      207760_s_atNCOR28180.6111.90E-032.00E-034.36E-021.22
      Fig. 2
      Fig. 2ROC curves and boxplots of the top four biomarker candidates involved in platinum + taxane resistance identified using the RFS at 6 months cohort. Only samples with serous histology and those treated with platinum and taxane combined therapy were included in the analysis.
      An independent analysis was performed using only patients with platinum monotherapy. Although there were 335 such patients, only 137 of these had RFS data, and only 109 of these were measured using the plus2 arrays. In this cohort, AKIP1 (p = 4.01E-03, AUC = 0.693), MARVELD1 (p = 1.96E-03, AUC = 0.708), AKIRIN2 (p = 4.48E-04, AUC = 0.735) and CFL1 (p = 8.02E-03, AUC = 0.667) reached significance while SERBP1, PDXK, TFE3, and NCOR2 were not significant.

      3.4 Validation

      The top potential biomarker candidates from the in-silico analysis were selected for further validation by RT-PCR in the OOI cohort of ovarian cancer patients. Table 4 lists the primer sequences used in the PCR validation. From the 81 fresh-frozen ovarian tissue samples, we excluded 15 samples due to insufficient follow-up data (n = 7), lack of chemotherapy (n = 1) and conflicting histological diagnosis (n = 8). From the remaining 66 samples, 47 samples were categorized as responders, and 19 samples were categorized as non-responders based on the RFS duration as described in the validation cohorts. All patients in the validation set had cancer of the serous subtype. The clinical characteristics of the specimens are presented in Fig. 1C.
      Table 4Quantitative PCR primers for the selected and reference genes.
      Gene symbolNCBI nucleotide sequence IDPrimer sequence (5′ → 3′)Length (bp)
      TFE3NM_006521.5F:GCTCCGAATTCAGGAACTAGAAC102
      R:CTGTCAGAAGCCGAAGTCGT
      NCOR2NM_006312.5F:CCACCCTCTGTCTCCTCAGT122
      R:AGGGGGTTGTAGGGGAATGG
      PDXKNM_003681.4F:ATCCAGTGTGCAAAAGCCCA186
      R:CAGGGACAAACACGGAGACA
      AKIP1NM_020642.3F:TTCTGTCACTGTGGGCTCAA85
      R:GAAGACCAGGTCCACGCTTT
      MARVELD1NM_031484.3F:GGGCCTGTAAGGTTTCCATGT148
      R:CCCCTACTGCCAGTGAAGAC
      CFL1NM_005507.2F:AAGAAGCTGACAGGGATCAAGC138
      R:GCCAGAAGGGGCTCACAAA
      AKIRIN2NM_018064.3F:ACAGCCTGCTAGCTATGTTTCA182
      R:AACCAGTTGCTGCTGCCTAA
      SERBP1NM_001018067.1F:GCAGGACCGACAAGTCAAGT79
      R:GCATCCAGTTAAGCCAGAGC
      GAPDHNM_002046.6F:AAATCAAGTGGGGCGATGCT86
      R:CAAATGAGCCCCAGCCTTCT
      ACTBNM_001101.5F:CTGTGGCATCCACGAAACTA200
      R:AGTACTTGCGCTCAGGAGGA
      In the validation, six genes reached a statistically significant correlation with response. All biomarker candidate genes were overexpressed in the nonresponder phenotype. These were TFE3 (p = 0.012, AUC = 0.718), NCOR2 (p = 0.048, AUC = 0.671), PDXK (p = 0.019, AUC = 0.702), AKIP1 (p = 0.002, AUC = 0.773), MARVELD1 (p = 0.044, AUC = 0.675) and AKIRIN2 (p = 0.042, AUC = 0.676). Detailed results for these genes are presented in Fig. 3.
      Fig. 3
      Fig. 3ROC curves and boxplots of RT-PCR-validated genes in the serous histology subtype clinical specimens. Only samples with serous histology and those treated with platinum and taxane combined therapy were retained for validation.
      The relative expression values in comparison to GAPDH and ACTB, including clinical information for each sample, are presented in Supplemental Table 1.

      3.5 Web application

      Finally, to enable the independent validation of our results and the analysis of novel gene candidates, the previously established ROC Plotter web application [
      • Fekete J.T.
      • Gyorffy B.
      ROCplot.org: validating predictive biomarkers of chemotherapy/hormonal therapy/anti-HER2 therapy using transcriptomic data of 3,104 breast cancer patients.
      ] was extended to include the ovarian cancer datasets described above. The registration-free web interface can be accessed at www.rocplot.org/ovar.

      4. Discussion

      Chemoresistance is a key problem in cancer treatment and is responsible for the poor prognosis of ovarian cancer patients. The identification of drug resistance-related genes enabling personalization of treatment selection is of utmost importance. The primary aim of this study was to identify potential predictive biomarkers that could predict the response to the most commonly used combination treatment, platinum and taxane, in serous ovarian tumors. Second, we aimed to validate these findings in an independent set of clinical specimens. Our final task was to extend our freely accessible online tool to enable the investigation of gene expression-based predictive biomarkers in ovarian cancer.
      Overall, we identified eight genes capable of classifying platinum and taxane drug responses. The independent validation results confirmed 6 of these genes. Some of these genes, namely, TFE3, NCOR2, PDXK and MARVELD1, were previously related to platinum- or taxane-based therapy resistance. Of these, TFE2, NCOR2, and PDXK were not significant in the platinum monotherapy cohort suggesting that these are markers linked to response to the combination therapy.
      The translocation of the transcription factor E3 (TFE3) with different fusion partners was described in renal carcinomas and alveolar soft part sarcomas [
      • Haq R.
      • Fisher D.E.
      Biology and clinical relevance of the micropthalmia family of transcription factors in human cancer.
      ]. It is overexpressed in head and neck squamous carcinoma treated with cisplatin-based chemotherapy. Higher expression of TFE3 indicated a poorer response to treatment [
      • Sun Z.J.
      • Yu G.T.
      • Huang C.F.
      • Bu L.L.
      • Liu J.F.
      • Ma S.R.
      • Zhang W.F.
      • Liu B.
      • Zhang L.
      Hypoxia induces TFE3 expression in head and neck squamous cell carcinoma.
      ]. Consistent with these findings, we observed elevated expression of TFE3 in the nonresponder cohort.
      The nuclear receptor corepressor 2 (NCOR2) is the repressor of the pregnane X receptor (PXR). PXR is a nuclear receptor that plays a role in the metabolism of different xenobiotics and endobiotics and was previously linked to cancer pathogenesis [
      • Pondugula S.R.
      • Mani S.
      Pregnane xenobiotic receptor in cancer pathogenesis and therapeutic response.
      ]. NCOR2-overexpressing head and neck cancer cell lines showed increased resistance to paclitaxel, cisplatin and 5-FU [
      • Rigalli J.P.
      • Reichel M.
      • Reuter T.
      • Tocchetti G.N.
      • Dyckhoff G.
      • Herold-Mende C.
      • Theile D.
      • Weiss J.
      The pregnane X receptor (PXR) and the nuclear receptor corepressor 2 (NCoR2) modulate cell growth in head and neck squamous cell carcinoma.
      ]. Our results also suggest that elevated expression of NCOR2 is one of the top biomarkers of resistance.
      Pyridoxal kinase (PDXK) is the key gene in the synthesis of pyridoxal-5-phosphate during B6 vitamin metabolism. Previous studies reported the key role of B6 vitamin in the uptake of cisplatin in A549 lung cancer cells, and high PDXK expression was associated with better disease outcome in lung cancer patients; however, the latter finding was unrelated to the patient's chemotherapy treatment [
      • Galluzzi L.
      • Marsili S.
      • Vitale I.
      • Senovilla L.
      • Michels J.
      • Garcia P.
      • Vacchelli E.
      • Chatelut E.
      • Castedo M.
      • Kroemer G.
      Vitamin B6 metabolism influences the intracellular accumulation of cisplatin.
      ]. In our clinical cohort, the elevated expression of PDXK was associated with a nonresponder phenotype, which seems to be a new feature of PDXK.
      The Marvel domain containing 1 (MARVELD1) protein is a member of the MARVEL domain containing proteins. These proteins are involved in cell cycle progression, chemotactic activity and endocytosis. Higher expression of MARVELD1 was associated with increased chemosensitivity to epirubicin and 10-hydroxycamptothecin in hepatocellular carcinoma cells [
      • Yu Y.
      • Zhang Y.
      • Hu J.
      • Zhang H.
      • Wang S.
      • Han F.
      • Yue L.
      • Qu Y.
      • Zhang Y.
      • Liang H.
      • Nie H.
      • Li Y.
      MARVELD1 inhibited cell proliferation and enhance chemosensitivity via increasing expression of p53 and p16 in hepatocellular carcinoma.
      ]. The inhibition of MARVELD1 repressed paclitaxel and cisplatin resistance in lung cancer cells [
      • Li J.
      • Yan H.
      • Zhao L.
      • Jia W.
      • Yang H.
      • Liu L.
      • Zhou X.
      • Miao P.
      • Sun X.
      • Song S.
      • Zhao X.
      • Liu J.
      • Huang G.
      Inhibition of SREBP increases gefitinib sensitivity in non-small cell lung cancer cells.
      ]. In contrast, elevated MARVELD1 expression inhibited arsenic-trioxide-induced apoptosis in liver cancer cells and was significantly related to worse overall survival of liver cancer patients [
      • Ma W.
      • Shen H.
      • Li Q.
      • Song H.
      • Guo Y.
      • Li F.
      • Zhou X.
      • Guo X.
      • Shi J.
      • Cui Q.
      • Xing J.
      • Deng J.
      • Yu Y.
      • Liu W.
      • Zhao H.
      MARVELD1 attenuates arsenic trioxide-induced apoptosis in liver cancer cells by inhibiting reactive oxygen species production.
      ]. Here, higher expression of this protein in ovarian cancer samples increased chemoresistance to platinum and taxane combination therapy.
      Two additional biomarker candidate genes (AKIP1 and AKIRIN2) have not been previously reported as potential resistance genes in platinum- and taxane-based chemotherapy. The A-kinase interacting protein 1 (AKIP1) is a nuclear protein that plays a role in NF-κB signaling [
      • Gao N.
      • Asamitsu K.
      • Hibi Y.
      • Ueno T.
      • Okamoto T.
      AKIP1 enhances NF-kappaB-dependent gene expression by promoting the nuclear retention and phosphorylation of p65.
      ]. Previously, higher expression of AKIP1 was described in breast cancer samples, and the higher expression correlated with worse survival [
      • Mo D.
      • Li X.
      • Li C.
      • Liang J.
      • Zeng T.
      • Su N.
      • Jiang Q.
      • Huang J.
      Overexpression of AKIP1 predicts poor prognosis of patients with breast carcinoma and promotes cancer metastasis through Akt/GSK-3beta/Snail pathway.
      ]. In another study, AKIP1 was identified as a regulator of WNT/β-catenin signaling activation, which promotes the metastatic relapse of hepatocellular carcinoma [
      • Cui Y.
      • Wu X.
      • Lin C.
      • Zhang X.
      • Ye L.
      • Ren L.
      • Chen M.
      • Yang M.
      • Li Y.
      • Li M.
      • Li J.
      • Guan J.
      • Song L.
      AKIP1 promotes early recurrence of hepatocellular carcinoma through activating the Wnt/beta-catenin/CBP signaling pathway.
      ]. Our results confirm the positive relationship between high AKIP1 expression and poor prognosis. Akirin 2 (AKIRIN2) is another nuclear protein that functions in B cell activation and humoral immune responses [
      • Tartey S.
      • Matsushita K.
      • Imamura T.
      • Wakabayashi A.
      • Ori D.
      • Mino T.
      • Takeuchi O.
      Essential function for the nuclear protein Akirin2 in B cell activation and humoral immune responses.
      ]. In a previous study, the gene was overexpressed in human cholangiocarcinoma cell lines and tumor tissues, and its elevated expression was associated with cell proliferation, migration, invasion, and angiogenesis [
      • Leng K.
      • Xu Y.
      • Kang P.
      • Qin W.
      • Cai H.
      • Wang H.
      • Ji D.
      • Jiang X.
      • Li J.
      • Li Z.
      • Huang L.
      • Zhong X.
      • Sun X.
      • Wang Z.
      • Cui Y.
      Akirin2 is modulated by miR-490-3p and facilitates angiogenesis in cholangiocarcinoma through the IL-6/STAT3/VEGFA signaling pathway.
      ]. AKIRIN2 knockdown led to decreased chemoresistance in temozolomide-treated glioblastoma cell lines [
      • Krossa S.
      • Schmitt A.D.
      • Hattermann K.
      • Fritsch J.
      • Scheidig A.J.
      • Mehdorn H.M.
      • Held-Feindt J.
      Down regulation of Akirin-2 increases chemosensitivity in human glioblastomas more efficiently than Twist-1.
      ]. Based on our study, the elevated expression of AKIRIN2 may also have an impact on the response to platinum-taxane combination therapy.
      The genes MARVELD1, AKIP1 and AKIRIN2 were also significant when performing the analysis in the platinum monotherapy treated cohort – these could be markers of resistance for both treatment settings. However, a limitation of this analysis is that only 109 patients were available for the platinum monotherapy cohort and an independent analysis in patients who received taxane only was not possible due to a very low sample number (n = 17).
      There are two notable limitations of the database utilized as a discovery set in our study: first, the number of patients is limited for some treatment cohorts, including those with targeted or second-line therapy. For this reason, we could only select the platinum-taxane combination for further validation. Second, the database contains incomplete clinical annotations for some of the samples. These limitations can be abridged by a future extension of the database.
      In summary, we collected 1816 ovarian cancer gene expression microarray samples with clinical information, including treatment and response data. Mining of this database has the potential to identify new predictive biomarkers, as we demonstrated here for platinum and taxane combination treatment. We validated a limited set of biomarker candidates by RT-PCR and identified six genes (TFE3, NCOR2, PDXK, AKIP1, MARVELD1 and AKIRIN2) with significant correlations with chemoresistance. The extended online analysis platform available at www.rocplot.org/ovar enables the discovery, validation and ranking of further predictive biomarker candidates in ovarian cancer.
      The following is the supplementary data related to this article.
      • Supplemental Table 1

        Clinical parameters and RT-PCR based expression of the validation group. Sample ID, histology of the tumor, stage, grade, debulk, response, chemotherapy (TXN/PLT means taxane + platinum) and the relative expression of the investigated genes.

      Acknowledgements

      The study was supported by the KH-129581 and the NVKP_16-1-2016-0037 grants of the National Research, Development and Innovation Fund , Hungary. The use of the computational infrastructure of Pázmány Péter University, provided within the National Bionics Program, is gratefully acknowledged. The authors acknowledge the support of ELIXIR Hungary (www.elixir-hungary.org).

      Author contributions

      Conceptualization: BGY, JTF; Clinical specimen collection: IP, IV, RNGy; Clinical specimen analysis: ÁŐ, JTF, Draft manuscript: JTF, BGY. Data interpretation and manuscript edition: JTF, BGY, IP, IV, RNGy, ÁŐ. All authors provided final approval of the manuscript.

      Declaration of competing interest

      The authors have declared no conflict of interest.

      References

        • Bray F.
        • Ferlay J.
        • Soerjomataram I.
        • Siegel R.L.
        • Torre L.A.
        • Jemal A.
        Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries.
        CA Cancer J. Clin. 2018; 68: 394-424
        • Menyhart O.
        • Fekete J.T.
        • Gyorffy B.
        Demographic shift disproportionately increases cancer burden in an aging nation: current and expected incidence and mortality in Hungary up to 2030.
        Clin. Epidemiol. 2018; 10: 1093-1108
        • Matulonis U.A.
        • Sood A.K.
        • Fallowfield L.
        • Howitt B.E.
        • Sehouli J.
        • Karlan B.Y.
        Ovarian cancer.
        Nat. Rev. Dis. Primers. 2016; 216061
        • Ozols R.F.
        Systemic therapy for ovarian cancer: current status and new treatments.
        Semin. Oncol. 2006; 33: S3-11
        • Crawford S.M.
        • Peace J.
        Does the nadir CA125 concentration predict a long-term outcome after chemotherapy for carcinoma of the ovary?.
        Ann. Oncol. 2005; 16: 47-50
        • Samimi G.
        • Varki N.M.
        • Wilczynski S.
        • Safaei R.
        • Alberts D.S.
        • Howell S.B.
        Increase in expression of the copper transporter ATP7A during platinum drug-based treatment is associated with poor survival in ovarian cancer patients.
        Clin. Cancer Res. 2003; 9: 5853-5859
        • Pljesa-Ercegovac M.
        • Savic-Radojevic A.
        • Matic M.
        • Coric V.
        • Djukic T.
        • Radic T.
        • Simic T.
        Glutathione Transferases: potential targets to overcome chemoresistance in solid tumors.
        Int. J. Mol. Sci. 2018; 19
        • Wang J.
        • Wu G.S.
        Role of autophagy in cisplatin resistance in ovarian cancer cells.
        J. Biol. Chem. 2014; 289: 17163-17173
        • Penzvalto Z.
        • Surowiak P.
        • Gyorffy B.
        Biomarkers for systemic therapy in ovarian cancer.
        Curr. Cancer Drug Targets. 2014; 14: 259-273
        • Nanayakkara A.K.
        • Follit C.A.
        • Chen G.
        • Williams N.S.
        • Vogel P.D.
        • Wise J.G.
        Targeted inhibitors of P-glycoprotein increase chemotherapeutic-induced mortality of multidrug resistant tumor cells.
        Sci. Rep. 2018; 8: 967
        • Orr G.A.
        • Verdier-Pinard P.
        • McDaid H.
        • Horwitz S.B.
        Mechanisms of Taxol resistance related to microtubules.
        Oncogene. 2003; 22: 7280-7295
        • Kampan N.C.
        • Madondo M.T.
        • McNally O.M.
        • Quinn M.
        • Plebanski M.
        Paclitaxel and its evolving role in the management of ovarian cancer.
        Biomed. Res. Int. 2015; 2015: 413076
        • Hsu K.F.
        • Shen M.R.
        • Huang Y.F.
        • Cheng Y.M.
        • Lin S.H.
        • Chow N.H.
        • Cheng S.W.
        • Chou C.Y.
        • Ho C.L.
        Overexpression of the RNA-binding proteins Lin28B and IGF2BP3 (IMP3) is associated with chemoresistance and poor disease outcome in ovarian cancer.
        Br. J. Cancer. 2015; 113: 414-424
        • Cheng W.F.
        • Huang C.Y.
        • Chang M.C.
        • Hu Y.H.
        • Chiang Y.C.
        • Chen Y.L.
        • Hsieh C.Y.
        • Chen C.A.
        High mesothelin correlates with chemoresistance and poor survival in epithelial ovarian carcinoma.
        Br. J. Cancer. 2009; 100: 1144-1153
        • Januchowski R.
        • Swierczewska M.
        • Sterzynska K.
        • Wojtowicz K.
        • Nowicki M.
        • Zabel M.
        Increased expression of several collagen genes is associated with drug resistance in ovarian cancer cell lines.
        J. Cancer. 2016; 7: 1295-1310
        • Chen C.
        • Chang Y.C.
        • Lan M.S.
        • Breslin M.
        Leptin stimulates ovarian cancer cell growth and inhibits apoptosis by increasing cyclin D1 and Mcl-1 expression via the activation of the MEK/ERK1/2 and PI3K/Akt signaling pathways.
        Int. J. Oncol. 2013; 42: 1113-1119
        • Lane D.
        • Matte I.
        • Garde-Granger P.
        • Laplante C.
        • Carignan A.
        • Rancourt C.
        • Piche A.
        Inflammation-regulating factors in ascites as predictive biomarkers of drug resistance and progression-free survival in serous epithelial ovarian cancers.
        BMC Cancer. 2015; 15: 492
        • Yu R.
        • Jin H.
        • Jin C.
        • Huang X.
        • Lin J.
        • Teng Y.
        Inhibition of the CSF-1 receptor sensitizes ovarian cancer cells to cisplatin.
        Cell Biochem. Funct. 2018; 36: 80-87
        • Yang S.D.
        • Ahn S.H.
        • Kim J.I.
        3-Oxoacid CoA transferase 1 as a therapeutic target gene for cisplatin-resistant ovarian cancer.
        Oncol. Lett. 2018; 15: 2611-2618
        • Sanchez-Alvarez R.
        • Martinez-Outschoorn U.E.
        • Lin Z.
        • Lamb R.
        • Hulit J.
        • Howell A.
        • Sotgia F.
        • Rubin E.
        • Lisanti M.P.
        Ethanol exposure induces the cancer-associated fibroblast phenotype and lethal tumor metabolism: implications for breast cancer prevention.
        Cell Cycle. 2013; 12: 289-301
        • Gautier L.
        • Cope L.
        • Bolstad B.M.
        • Irizarry R.A.
        affy—analysis of Affymetrix GeneChip data at the probe level.
        Bioinformatics. 2004; 20: 307-315
        • Sims A.H.
        • Smethurst G.J.
        • Hey Y.
        • Okoniewski M.J.
        • Pepper S.D.
        • Howell A.
        • Miller C.J.
        • Clarke R.B.
        The removal of multiplicative, systematic bias allows integration of breast cancer gene expression datasets - improving meta-analysis and prediction of prognosis.
        BMC Med. Genet. 2008; 1: 42
        • Fekete J.T.
        • Gyorffy B.
        ROCplot.org: validating predictive biomarkers of chemotherapy/hormonal therapy/anti-HER2 therapy using transcriptomic data of 3,104 breast cancer patients.
        Int. J. Cancer. 2019; 145: 3140-3151
        • Haq R.
        • Fisher D.E.
        Biology and clinical relevance of the micropthalmia family of transcription factors in human cancer.
        J. Clin. Oncol. 2011; 29: 3474-3482
        • Sun Z.J.
        • Yu G.T.
        • Huang C.F.
        • Bu L.L.
        • Liu J.F.
        • Ma S.R.
        • Zhang W.F.
        • Liu B.
        • Zhang L.
        Hypoxia induces TFE3 expression in head and neck squamous cell carcinoma.
        Oncotarget. 2016; 7: 11651-11663
        • Pondugula S.R.
        • Mani S.
        Pregnane xenobiotic receptor in cancer pathogenesis and therapeutic response.
        Cancer Lett. 2013; 328: 1-9
        • Rigalli J.P.
        • Reichel M.
        • Reuter T.
        • Tocchetti G.N.
        • Dyckhoff G.
        • Herold-Mende C.
        • Theile D.
        • Weiss J.
        The pregnane X receptor (PXR) and the nuclear receptor corepressor 2 (NCoR2) modulate cell growth in head and neck squamous cell carcinoma.
        PLoS One. 2018; 13e0193242
        • Galluzzi L.
        • Marsili S.
        • Vitale I.
        • Senovilla L.
        • Michels J.
        • Garcia P.
        • Vacchelli E.
        • Chatelut E.
        • Castedo M.
        • Kroemer G.
        Vitamin B6 metabolism influences the intracellular accumulation of cisplatin.
        Cell Cycle. 2013; 12: 417-421
        • Yu Y.
        • Zhang Y.
        • Hu J.
        • Zhang H.
        • Wang S.
        • Han F.
        • Yue L.
        • Qu Y.
        • Zhang Y.
        • Liang H.
        • Nie H.
        • Li Y.
        MARVELD1 inhibited cell proliferation and enhance chemosensitivity via increasing expression of p53 and p16 in hepatocellular carcinoma.
        Cancer Sci. 2012; 103: 716-722
        • Li J.
        • Yan H.
        • Zhao L.
        • Jia W.
        • Yang H.
        • Liu L.
        • Zhou X.
        • Miao P.
        • Sun X.
        • Song S.
        • Zhao X.
        • Liu J.
        • Huang G.
        Inhibition of SREBP increases gefitinib sensitivity in non-small cell lung cancer cells.
        Oncotarget. 2016; 7: 52392-52403
        • Ma W.
        • Shen H.
        • Li Q.
        • Song H.
        • Guo Y.
        • Li F.
        • Zhou X.
        • Guo X.
        • Shi J.
        • Cui Q.
        • Xing J.
        • Deng J.
        • Yu Y.
        • Liu W.
        • Zhao H.
        MARVELD1 attenuates arsenic trioxide-induced apoptosis in liver cancer cells by inhibiting reactive oxygen species production.
        Ann. Transl. Med. 2019; 7: 200
        • Gao N.
        • Asamitsu K.
        • Hibi Y.
        • Ueno T.
        • Okamoto T.
        AKIP1 enhances NF-kappaB-dependent gene expression by promoting the nuclear retention and phosphorylation of p65.
        J. Biol. Chem. 2008; 283: 7834-7843
        • Mo D.
        • Li X.
        • Li C.
        • Liang J.
        • Zeng T.
        • Su N.
        • Jiang Q.
        • Huang J.
        Overexpression of AKIP1 predicts poor prognosis of patients with breast carcinoma and promotes cancer metastasis through Akt/GSK-3beta/Snail pathway.
        Am. J. Transl. Res. 2016; 8: 4951-4959
        • Cui Y.
        • Wu X.
        • Lin C.
        • Zhang X.
        • Ye L.
        • Ren L.
        • Chen M.
        • Yang M.
        • Li Y.
        • Li M.
        • Li J.
        • Guan J.
        • Song L.
        AKIP1 promotes early recurrence of hepatocellular carcinoma through activating the Wnt/beta-catenin/CBP signaling pathway.
        Oncogene. 2019; 38: 5516-5529
        • Tartey S.
        • Matsushita K.
        • Imamura T.
        • Wakabayashi A.
        • Ori D.
        • Mino T.
        • Takeuchi O.
        Essential function for the nuclear protein Akirin2 in B cell activation and humoral immune responses.
        J. Immunol. 2015; 195: 519-527
        • Leng K.
        • Xu Y.
        • Kang P.
        • Qin W.
        • Cai H.
        • Wang H.
        • Ji D.
        • Jiang X.
        • Li J.
        • Li Z.
        • Huang L.
        • Zhong X.
        • Sun X.
        • Wang Z.
        • Cui Y.
        Akirin2 is modulated by miR-490-3p and facilitates angiogenesis in cholangiocarcinoma through the IL-6/STAT3/VEGFA signaling pathway.
        Cell Death Dis. 2019; 10: 262
        • Krossa S.
        • Schmitt A.D.
        • Hattermann K.
        • Fritsch J.
        • Scheidig A.J.
        • Mehdorn H.M.
        • Held-Feindt J.
        Down regulation of Akirin-2 increases chemosensitivity in human glioblastomas more efficiently than Twist-1.
        Oncotarget. 2015; 6: 21029-21045