Unlike high-grade SAC, ovarian CCC are generally p53 wild type and have a lower frequency of breast cancer 1 (BRCA1) and BRCA2 mutations [1415]. Table 1 summarizes the critical genetic alterations observed in ovarian CCC [151617181920212223]. The most frequent and important alterations are AT rich interactive domain 1A (ARID1A) and phosphatidylinositol-45-bisphosphate 3-kinase catalytic subunit α (PIK3CA) mutations [1617]. It has been reported that tumors with ARID1A mutations also frequently harbor phosphatase and tensin homolog (PTEN) or PIK3CA mutations, suggesting their collaboration in CCC tumorigenesis [24]. Consistent with this, recent investigations involving genetically engineered mouse model (GEM) demonstrated that inactivation of ARID1A alone is insufficient for tumor initiation; it requires additional genetic alterations such as PIK3CA to drive CCC tumorigenesis [25].

The protein expressions in ovarian CCC have been intensively investigated [232627282930313233343536]. Table 2 is listing the proteins highly expressed in ovarian CCC. Interestingly, these proteins can be classified into five pathways: phosphatidylinositol 3-kinase (PI3K)/AKT/mammalian target of rapamycin (mTOR) pathway, hypoxia-inducible factor 1α (HIF-1α)/vascular endothelial growth factor (VEGF) pathway, hepatocyte nuclear factor 1β (HNF-1β) pathway, interleukin 6 (IL-6)/signal transducer and activator of transcription 3 (STAT3) pathway and MET pathway. These pathways may be important targets for the treatment of ovarian CCC, and inhibitors targeting some of these pathways are in clinical development.

GEM has advantages over nude mice as an experimental model, because the tumors develop spontaneously, the site of tumors is orthotopic, and they are immuno-competent [37]. As shown in Table 3, although p53, PTEN, PIK3CA, and ARID1A are known to be key genes for ovarian carcinogenesis, none of the genes, when mutated individually, could lead to the development of ovarian cancer in mice. However, the combination of at least 2 mutations had resulted in the development of serous, endometrioid, or undifferentiated adenocarcinomas. Very recently, a mouse model of ovarian CCC carrying coexisting ARID1A and PIK3CA mutations has been introduced [25]. By using the intra-bursal gene-delivery method, the authors induced the ARID1A-loss and PIK3CA-hyperactivation, and succeeded in the development of ovarian CCC. Importantly, consistent with the finding from ovarian CCC in human, PI3K/AKT/mTOR pathway is hyperactivated in ovarian CCC in this mouse model. When the mice were treated with a PI3K inhibitor, the survival of mice was significantly prolonged [25]. Additional important finding in this mouse model is the activation of IL-6/STAT3 signaling, which is also consistent with the findings in human CCC [3536]. The precise mechanisms responsible for the IL-6/STAT3 signaling activation remain unknown, but the authors speculated that ARID1A may negatively regulate the expression of IL-6, and oncogenic PIK3CA mutation enhances the IL-6 production in the absence of negative-regulation by ARID1A in ovarian CCC [25]. We believe that the mouse model of ovarian CCC can be used for the response assessment of the candidate drugs, biomarker analyses, mechanistic investigations for the understanding of chemoresistance or carcinogenesis, or chemoprevention study, that can lead to the acceleration of the development of effective therapies against ovarian CCC.

HNF-1β is a transcriptional factor which is commonly (over 95%) expressed in ovarian CCC [31]. Previous investigations suggested that it regulates the expression of proteins/genes such as annexin A4, uridine diphosphate (UDP)-glucuronosyl transferase 1A1 (UGT1A1), osteopontin, or insulin-like growth factor-binding protein 1 (IGFBP-1) that are important for cancer progression [31]. However, the specific inhibitor of HNF-1β has not been developed yet, and the significance of HNF-1β as a therapeutic target remains to be elucidated. Because of its high expression rate in ovarian CCC [33], HNF-1β is now used as a diagnostic marker of ovarian CCC [38].

Both VEGF and IL-6 bind to their corresponding receptors, activate downstream effectors, and promote the progression of human cancers (Fig. 1).

VEGF promotes proliferation, survival, and migration of endothelial cells to induce tumor angiogenesis, and inhibitors of VEGF or VEGF receptor (VEGFR) have now approved in the treatment of ovarian cancer [39]. It was reported that VEGF is commonly (90% to 100%) expressed in both early-stage and advanced-stage ovarian CCC, and its expression is associated with patient’s survival [30].

IL-6 is a pleiotropic proinflammatory cytokine that has emerged as a mediator of pivotal processes such as cell proliferation, angiogenesis, and chemoresistance within tumor microenvironment. It has been reported that IL-6/STAT3 signaling is frequently activated in ovarian CCC, and IL-6 receptor (IL-6R) expression is associated with survival [3536]. Collectively, these results indicate that the progression of ovarian CCC is dependent, at least in part, on VEGF and IL-6/STAT3 pathways. Currently, the clinical activity of VEGFR-inhibitor sunitinib is being evaluated in a phase II trial in patients with ovarian CCC [40]. Although the final results have not been reported, the data presented in the Society of Gynecologic Oncology 2015 meeting has suggested that sunitinib has minimal clinical activity in patients with recurrent CCC as the second- or third-line treatment, with a response rate of 6.7% [41]. As for IL-6-targeted therapy, no clinical trials have been conducted yet.

The PI3K/AKT/mTOR pathway plays a critical role in the malignant transformation of human tumors and their subsequent growth, proliferation, and metastasis [42]. Previous studies have shown that ovarian CCC often exhibit genetic alterations in one or more components of the PI3K/AKT/mTOR signaling pathway (Table 1); PIK3CA mutation in 33% to 40% [17], PTEN mutation in 5% [20], or AKT2 amplification in 14% [23]. In addition to these genetic alterations, the loss of PTEN expression has been detected in 40% of ovarian CCC [43]. Consistent with these alterations, our group examined tissue microarrays of 98 primary ovarian tumors (52 CCC and 46 SAC) to tissue microarrays and demonstrated that AKT, mammalian target of rapamycin complex 1 (mTORC1), and mTORC2 are more frequently activated in CCC than SAC (CCC vs. SAC for AKT, 69.2% vs. 63%; mTORC1, 86.6% vs. 50%; and mTORC2, 71.2% vs. 45.7%) [262728]. Importantly, PI3K/AKT/mTOR inhibitors had marked anti-tumor effects in ovarian cancer cells exhibiting high AKT/mTORC1 activity, but minimal effects in ovarian cancer cells displaying low AKT/mTORC1 activity [262728]. Thus, PI3K/AKT/mTOR pathway is regarded as an attractive therapeutic target in in ovarian CCC.

The PI3K/AKT/mTOR pathway inhibitors include PI3K inhibitors, AKT inhibitors, mTORC1 inhibitors, and dual inhibitors (mTORC1/2 inhibitors or PI3K/mTOR inhibitors) (Fig. 2). These inhibitors are currently in various stages of clinical development in the setting of single agent or in combination with other targeted agents in ovarian cancer patients (Table 4). Of these, GOG-0268 is a phase II trial specifically targeting ovarian CCC, and is examining temsirolimus in combination with carboplatin and paclitaxel followed by temsirolimus consolidation as a first-line therapy for patients with stage III to IV diseases. As the dual inhibitors are theoretically more efficacious than the conventional inhibitors (mTORC1 inhibitor or PI3K inhibitors) [28], the clinical activity of dual inhibitors should be investigated in future trials in patients with ovarian CCC.

MET is a receptor tyrosine kinase that, after binding with hepatocyte growth factor, activates Raf/Ras/mitogen-activated protein kinases (MAPK) and PI3K/AKT/mTOR signaling pathways and stimulates cancer cell proliferation, migration, and invasion (Fig. 1). Recent investigations have suggested that MET overexpression and gene amplification were commonly observed in CCC with their frequencies of 22% and 24% to 37%, respectively [2344]. Moreover, both MET overexpression and gene amplification were shown to be associated with worse prognoses in CCC patients [2545]. Furthermore, MET knockdown in CCC cell lines with MET amplification resulted in both increased apoptosis and senescence in vitro [23]. These results suggest that MET-targeted therapy may have therapeutic efficacy in patients with CCC. Currently, the safety and the activity of MET inhibitor (INC280) are being evaluated in a phase I trial in patients with advanced MET-dependent solid tumor including ovarian cancer [45].

Poly (ADP-ribose) polymerase (PARP) is a protein that is involved in the repair of single-strand breaks (SSBs) of DNA. When PARP is inhibited, unrepaired SSBs can result in double-strand breaks (DSBs). As DSBs of DNA is repaired mainly by homologous recombination pathway (HR), DSBs induced by PARP-inhibition can lead to lethal DNA damages in HR-deficient cells: i.e., cells with BRCA1/2 mutations (Fig. 3). Consistent with these findings, ovarian cancers occurring in patients with deleterious germline mutations in BRCA1 or BRCA2 have shown significant sensitivity to PARP inhibitors [4647]. It is also becoming clear that germline BRCA1/2 mutations are neither necessary nor sufficient for patients to derive benefit from PARP inhibitors. Several mechanisms that induce HR dysfunction have been identified: i.e., BRCA1 promoter methylation [48], or loss or reduction in proteins necessary for HR. PTEN loss may be a common contributing event causing HR dysfunction [49], and the increased PARP inhibitor susceptibility was demonstrated in a series of cell lines with PTEN mutation [49]. Moreover, previous investigations have demonstrated that PI3K inhibition is associated with HR dysfunction, leading to increased DNA damage and more sensitization of cancer cells to PARP inhibitors [50]. Thus, although the germline mutations in BRCA1 or BRCA2 are observed only in 6% of ovarian CCC patients [15], we consider that the activity of PARP inhibitor plus PI3K/AKT/mTOR inhibitor is worth investigating in this patient population, in which loss of PTEN expression is observed in 40% of cases [43]. Currently, phase I trials of combined treatment involving the PARP inhibitor olaparib and pan-PI3K inhibitors or AKT-inhibitor are enrolling cancer patients including ovarian cancer (Table 4).

Nucleotide excision repair (NER) is a versatile DNA repair system that recognizes and acts against DNA damage induced by platinum-based agents [51]. According to previous reports, increased NER activity is associated with resistance to platinum-based agents, and NER deficiency is associated with increased sensitivity to platinum-based agents [5253]. Previous study of ovarian cancer showed that NER activity is increased in ovarian CCC than in any other histological subtypes of ovarian cancer [54], indicating that the increased NER activity may be associated with, at least in part, the platinum resistance in ovarian CCC. Thus, it would be reasonable to treat ovarian CCC using agents whose effects are not subject to NER repair.

Trabectedin, an anticancer agent, has recently become the focus of attention for researchers investigating the treatment of ovarian cancer. Trabectedin is an antineoplastic agent that was originally derived from the Caribbean marine tunicate Ecteinascidia turbinata. It binds covalently to the minor groove of DNA, bending the DNA toward the major groove and disrupting its transcription, leading to G₂-M cell-cycle arrest and ultimately apoptosis [55]. On the basis of the positive results of a phase III clinical study (OVA301 study) [56], the use of trabectedin in combination with pegylated liposomal doxorubicin was approved by the European Union in 2009 as a treatment for relapsed platinum-sensitive ovarian cancer.

Trabectedin interacts with the NER machinery in an unusual manner. An elegant study demonstrated that NER-deficient cells (deficient in NER-related genes) exhibited resistance to trabectedin and that their sensitivity to trabectedin was restored by the transfection of the corresponding genes [57]. These findings are in clear contrast with the results obtained for platinum-based agents. Moreover, interestingly, a previous study demonstrated that cells in the G₁ phase are most sensitive to trabectedin [58]. As NER activity is increased in ovarian CCC cells and more than half of ovarian CCC cells are in G1-phase [59], trabectedin might be effective in patients with ovarian CCC. In preclinical investigations, we found that trabectedin shows the greatest in vitro antiproliferative effect than any other anticancer agents that are clinically used in the treatment of ovarian CCC [59]. We also found that the in vitro antiproliferative effect of trabectedin is further enhanced by the co-treatment with mTORC1 inhibitor [59]. Moreover, a strong synergy was observed when trabectedin was combined with SN-38, an active metabolite of irinotecan [60]. These results indicate that trabectedin plus mTORC1 inhibitor and/or irinotecan is worth investigating in the future clinical trials [60]. Currently, the clinical activity of trabectedin plus mTORC1 inhibitor everolimus is being evaluated in a phase II study involving recurrent ovarian CCC patients in Japan, and a preliminary evaluation showed that the drug exhibited significant activity, with a response rate of 14.3% and a clinical benefit rate of 42.9% [61].

Lurbinectedin (PM01183) is a novel synthetic agent derived from trabectedin. It is also a covalent DNA minor groove binder and is structurally similar to trabectedin except for the replacement of a tetrahydroisoquinoline present in trabectedin by a tetrahydro β-carboline. This structural difference confers pharmacokinetic and pharmacodynamic benefits, which result in decreased toxicities and allow treatment regimens with increased dose intensity that lead to increased antitumor activity when compared to trabectedin [62]. In preclinical studies, Lurbinectedin exhibited broad antitumor activity against human cancer cell lines in vitro [63]. Lurbinectedin also significantly inhibited the growth of a wide variety of human cancer xenografts in athymic mice [62]. Following the encouraging results obtained in preclinical studies and phase I to II clinical trials [626364], a phase III trial investigating the activity of lurbinectedin versus pegylated liposomal doxorubicin or topotecan against recurrent ovarian cancer is currently being conducted [45]. We hope this agent be evaluated in the future clinical trials in patients with ovarian CCC.