The mesenchymal transition subtype more responsive to dose dense taxane chemotherapy combined with carboplatin than to conventional taxane and carboplatin chemotherapy in high grade serous ovarian carcinoma: A survey of Japanese Gynecologic Oncology Group study (JGOG3016A1)

https://doi.org/10.1016/j.ygyno.2019.02.010Get rights and content

Highlights

  • The morphologically defined mesenchymal transition type is highly representative of the mesenchymal transcriptome subtype.

  • The mesenchymal transition type has the worst prognosis and shows an increased rate of suboptimal surgery.

  • The mesenchymal transition type benefits from dose dense taxane and carboplatin chemotherapy compared to conventional TC.

Abstract

Objective

Recently, we established new histopathological subtypes of high-grade serous ovarian cancer (HGSOC) that include the mesenchymal transition (MT) type, the immune reactive (IR) type, the solid and proliferative (SP) type and the papillo-glandular (PG) type. Furthermore, we identified that the mesenchymal transcriptome subtype might be sensitive to taxane. We investigated whether these different histopathological subtypes of HGSOC require individualized chemotherapy for optimal treatment.

Methods

We conducted the Japanese Gynecologic Oncology Group (JGOG) 3016A1 study, wherein we collected hematoxylin and eosin slides (total n = 201) and performed a histopathological analysis of patients with HGSOC registered in the JGOG3016 study, which compared the efficacy of conventional paclitaxel and carboplatin (TC) and dose-dense TC (ddTC). We analyzed the differences in progression-free survival (PFS) and overall survival (OS) among the four histopathological subtypes. We then compared the PFS between the TC group and the ddTC group for each histopathological subtype.

Results

There were significant differences in both PFS and OS among the four histopathological subtypes (p = 0.001 and p < 0.001, respectively). Overall, the MT subtype had the shortest PFS (median 1.4 y) and OS (median 3.6 y). In addition, the MT subtype had a longer PFS in the ddTC group (median 1.8 y) than in the TC group (median 1.2 y) (p = 0.01). Conversely, the other types had no significant difference in PFS when the two regimens were compared.

Conclusions

The MT type of HGSOC is sensitive to taxane; therefore, the ddTC regimen is recommended for this histopathological subtype.

Introduction

Ovarian cancer is the leading cause of death among gynecologic malignancies [1]. High-grade serous ovarian carcinoma (HGSOC) is the most frequent histopathological type of ovarian cancer, accounting for approximately 70% of ovarian cancers in the US [2] and 33–36% in Japan [3] [4]. The majority of HGSOCs are diagnosed at advanced stages; therefore, pharmacological treatment is necessary for this disease.

Paclitaxel and carboplatin combination (TC) therapy has been used as a standard first-line chemotherapy against ovarian cancer for two decades [5,6]. The Japanese Gynecologic Oncology Group (JGOG) conducted the JGOG3016 study, a randomized controlled trial that compared dose dense weekly paclitaxel combined with tri-weekly carboplatin (ddTC) and conventional tri-weekly paclitaxel and carboplatin (TC) treatment in patients with stage II-IV ovarian cancer, tubal cancer, and peritoneal cancer. That study demonstrated that progression-free survival (PFS) was significantly prolonged in patients taking ddTC than in those taking conventional TC as a first-line chemotherapy regimen [7]. However, the sensitivity to chemotherapy varies widely among cancer patients, and it is now considered necessary that individualized treatments should be administered in a clinical setting [8,9].

The Cancer Genome Atlas (TCGA) project revealed that HGSOCs are divided into four transcriptome subtypes: mesenchymal, immunoreactive, proliferative and differentiated [10,11]. Among these, the mesenchymal type has the worst prognoses, whereas the immunoreactive type has the best prognoses [9,11]. It is widely accepted that the gene expression profile of cancer is strongly related to the characteristics in its biology and its sensitivity to treatment. However, ovarian cancer subtypes with their unique gene expression have yet to be investigated for drug treatments in clinical settings [12]. Four obstacles to performing gene expression microarray analyses include cost, maintaining uniform conditions due to the variety of experimental procedures, inconsistent quality of nucleotides and the batch effect, which is due to systematic variation between groups of samples (batches) resulting from experimental features that are not of biological interest [13].

Previous studies showed that a remarkable desmoplastic reaction occurred in the mesenchymal subtypes and a marked T lymphocyte infiltration into the tumor was seen in the immunoreactive subtypes [14]. These reports suggested that in the HGSOC, the tumor microenvironment, specifically the interaction of the stroma and tumor, affected the gene expression profile and prognoses of the patients [14,15]. From this idea, we established a novel histopathological classification based on the tumor and stromal findings, corresponding to these four transcriptome subtypes of HGSOC. Of the four subtypes—mesenchymal transition (MT) type, immune reactive (IR) type, solid and proliferative (SP) type and papilloglandular (PG) type—the MT type has the worst prognoses and the IR type has the best prognoses [16].

Furthermore, we discovered that the mesenchymal transcriptome subtype could be sensitive to taxane. Therefore, we hypothesized that the mesenchymal transition (MT) type which represents the “mesenchymal” transcriptome subtype might be sensitive to taxane. [16,17] If this histopathological classification can be used to assess a subtype's drug sensitivity, individualized treatments can be developed based on hematoxylin and eosin (HE) slide analyses. This would offer a low cost and feasible approach to treatment. Subsequently, we conducted the JGOG3016A1 study, in which HGSOC cases registered in the JGOG3016 study were classified into histopathological subtypes. Consequently, a subtype more responsive to the ddTC regimen than to the TC regimen was identified. This study would be helpful to develop an individualized treatment of ovarian cancer.

Section snippets

Gene expression microarray analysis

The detailed methods of gene expression microarray analysis are described in the supplementary text. Briefly, we calculated prediction scores for taxane sensitivity (T-scores) [17] and the mesenchymal gene expression subtype score (CLOVAR_MES_UP_score) [11] within the three HGSOC gene expression microarray datasets; TCGA [10], GSE9891 [14], and KOV, which are our previously reported data set [16,18,19].

JGOG3016A1 study

Following the microarray-based prediction, we performed a study referred to as JGOG3016A1.

Results

First, we examined the correlation between T-scores representing taxane predictive scores [17] and CLOVER_MES_UP-scores [11] using the mesenchymal gene expression subtype in the HGSOC gene expression microarray datasets. In the three HGSOC datasets, TCGA, GSE9891 and KOV, these scores showed a strong positive correlation (Supplementary Fig. S1A; r = 0.87, p < 0.001, Supplementary Fig. S1B; r = 0.88, p < 0.001, Supplementary Fig. S1C; r = 0.92, p < 0.001, respectively). We found that T-scores

Discussion

For HGSOC, the CLOVER_MES_UP score that represents the Mesenchymal gene expression subtype [11] and T-score that represents taxane sensitivity [17] showed strong positive correlation (Supplementary Fig. S1A, 1B, 1C). As expected from these data, the MT histopathological type, that correlated with the mesenchymal gene expression subtype [16], had higher T-scores than other histopathological subtypes (Supplementary Fig. S1D), indicating that the MT type may be taxane sensitive. So far,

Conclusion

The MT subtype, which tends to result in suboptimal surgery and has the worst prognoses of all the subtypes in HGSOC, benefits more greatly from the ddTC regimen compared to the conventional TC one. We believe that the pathological diagnosis for the MT subtype in HGSOC has the potential to lead to their greater individualization. Further clinical validation study to investigate the reproducibility and basic research to discover the mechanism will be required as a future plan.

Acknowledgments of research support for the study

We would like to extend our appreciation to all group members related to this study.

Japanese Gynecologic Oncology Group (JGOG)

Tabata Tsutomu, Shoji Kamiura, Keiichi Fujiwara, Junzo Kigawa, Kimiyo Ushijima, Kentaro Kai, Rina Kato, Tatsuru Ohara, Hiroyuki Fujita, Hidemichi Watari, Hideki Tokunaga, Shigeki Niimi, Fumiaki Suzuki, Iemasa Kou, Eizo Kimura, Kohei Kotera, Ayumu Matsuoka, Yasuyuki Hirashima, Hiroyuki Fujiwara, Kotaro Sueoka, Satomi Aihara, Shin Nishio, Hiroshi Makino, Masashi Moriyama,

Author contributions

Planning and execution of this survey: RM, NM, AO, IK, MM; Sample preparation: HT, MY, HI, MS, TN, OT, MT, TS, TS, SI, KT, HN; Data analysis: RM, HM; Manuscript authoring: RM, NM, MM; Contextualization of findings of this work: RM, MN, HM, MM. All authors read and approved of the final manuscript.

Conflict of interest statement

Noriomi Matsumura reports that this work was supported by JSPS KAKENHI Grant Number 26253080.

Toru Sugiyama has an honoraria with BMS.

The other authors have no conflict of interest to declare.

References (39)

  • W. Yamagami et al.

    Clinical statistics of gynecologic cancers in Japan

    J. Gynecol. Oncol.

    (2017)
  • R.F. Ozols et al.

    Phase III trial of carboplatin and paclitaxel compared with cisplatin and paclitaxel in patients with optimally resected stage III ovarian cancer: a Gynecologic Oncology Group study

    J. Clin. Oncol. Off. J. Am. Soc. Clin. Oncol.

    (2003)
  • W.P. McGuire et al.

    Cyclophosphamide and cisplatin versus paclitaxel and cisplatin: a phase III randomized trial in patients with suboptimal stage III/IV ovarian cancer (from the Gynecologic Oncology Group)

    Semin. Oncol.

    (1996)
  • L. Mittempergher

    Genomic characterization of high-grade serous ovarian cancer: dissecting its molecular heterogeneity as a road towards effective therapeutic strategies

    Curr. Oncol. Rep.

    (2016)
  • G.E. Konecny et al.

    Prognostic and therapeutic relevance of molecular subtypes in high-grade serous ovarian cancer

    J. Natl. Cancer Inst.

    (2014)
  • Cancer Genome Atlas Research N

    Integrated genomic analyses of ovarian carcinoma

    Nature

    (2011)
  • R.G. Verhaak et al.

    Prognostically relevant gene signatures of high-grade serous ovarian carcinoma

    J. Clin. Investig.

    (2013)
  • D.D. Bowtell et al.

    Rethinking ovarian cancer II: reducing mortality from high-grade serous ovarian cancer

    Nat. Rev. Cancer

    (2015)
  • S.E. Reese et al.

    A new statistic for identifying batch effects in high-throughput genomic data that uses guided principal component analysis

    Bioinformatics

    (2013)
  • Cited by (15)

    • Magnetic Resonance Imaging and Diffusion Weighted Imaging-Based Histogram in Predicting Mesenchymal Transition High-Grade Serous Ovarian Cancer

      2023, Academic Radiology
      Citation Excerpt :

      Patients with mesenchymal subtype were reported had a median survival period of 23 months and platinum resistance rate of 63%, while those with non-mesenchymal subtype had a median survival period of 46 months and platinum resistance rate of 23%, respectively (4, 8, 16). Early non-invasive prediction of MT HGSOC could provide effective treatment guidance to assist clinicians in tumor management (6, 8). This study found that the risk factors of indiscrete primary ovarian mass, lower mean ADC value, and lower median ADC value were independently associated with MT HGSOC, and that the combination of all criteria showed satisfactory predictive performance with sensitivity greater than 75%.

    • Impact of veliparib, paclitaxel dosing regimen, and germline BRCA status on the primary treatment of serous ovarian cancer – an ancillary data analysis of the VELIA trial

      2022, Gynecologic Oncology
      Citation Excerpt :

      Because previous clinical trials of DD versus Q3W paclitaxel dosing for ovarian cancer did not report outcomes according to biomarker status [7–9], further studies are warranted to understand if specific subgroups of patients will benefit from DD paclitaxel dosing, particularly those with platinum-resistant and/or HRP ovarian cancer. JGOG2016A1, which was a survey study of JGOG3016, explored the efficacy of DD paclitaxel by histological subtype of high-grade serous carcinoma, and identified subgroups that were sensitive to DD regimen and subgroups that were less sensitive [31]. The correlation between these histological subtypes and BRCA/HRD status has not yet been investigated.

    View all citing articles on Scopus
    View full text