Histotype and molecular subtype integration in epithelial ovarian cancer: implications for targeted treatment.

The histotype-specific molecular landscapes and therapeutic evidence reviewed above support an integrated framework for therapeutic stratification in EOC. Across histotypes, recurrent genomic alterations, dysregulated signaling pathways, and clinically relevant protein-expression patterns define distinct molecular alterations with potential therapeutic implications. However, the presence of a molecular alteration alone does not establish therapeutic actionability. Clinical translation requires integration of histologic context, appropriate biomarker testing, matched therapeutic strategies, and the maturity of supporting evidence. Accordingly, we summarized these elements in a unified histotype-molecular therapeutic stratification framework that links histologic classification, biomarker assessment, treatment selection, and evidence level (Table 6 ). Table 6 Integrated histotype-molecular framework for therapeutic stratification in epithelial ovarian cancer Histotype Actionable alteration / biomarker Required assay Agent Evidence level Key caveat HGSOC BRCA1/2 mutation Germline and/or somatic BRCA testing; tumor NGS panel Olaparib; Niraparib; Rucaparib Approved; setting- and region-specific Regulatory indications differ by agent, line of therapy, maintenance versus treatment setting, and region; germline and somatic events should be distinguished. HRD-positive Validated HRD assay ± BRCA/NGS testing Olaparib + Bevacizumab; Niraparib Approved; indication-specific HRD definitions and thresholds vary across assays; HRD should not be equated with BRCA mutation alone. FRα-positive FRα/FOLR1 IHC using an approved companion diagnostic Mirvetuximab soravtansine Approved Approved for FRα-positive platinum-resistant epithelial ovarian, fallopian tube, or primary peritoneal cancer after 1-3 prior systemic regimens; not histotype-restricted. TP53 Y220C mutation Tumor NGS panel Rezatapopt (PC14586) Phase I; Phase II ongoing Applies only to the rare TP53 Y220C subset; evidence remains early-phase and is not yet histotype-specific. TP53 mutation, ATM alteration, BRCA-wild-type or replication-stress phenotype Tumor NGS panel ± IHC Berzosertib (ATRi); Adavosertib (WEE1i); Prexasertib (CHK1i) Phase II No standardized predictive biomarker is clinically validated for replication-stress and ATM/TP53 pathway defects; TP53 mutation alone does not establish WEE1 inhibitor sensitivity CCNE1 amplification Copy-number profiling by NGS, SNP array, or FISH Adavosertib (WEE1i); ATRi + WEE1i Phase II for WEE1i and preclinical for combination checkpoint inhibition CCNE1 amplification is a potential biomarker, but copy-number threshold and prospective clinical utility require further validation. LGSOC KRAS mutation Tumor NGS panel or validated KRAS mutation test Avutometinib + Defactinib Approved Approval is specifically for KRAS-mutated recurrent LGSOC after prior systemic therapy; continued approval may depend on confirmatory benefit from RAMP-301. MAPK pathway alterations except for KRAS mutation Tumor NGS panel Trametinib; Binimetinib; Selumetinib Phase II to phase II/III Trametinib has randomized phase II/III evidence; binimetinib and selumetinib data are limited by early termination, single-arm design, or post hoc interpretation. BRAF V600 mutation Tumor NGS panel Vemurafenib Phase II basket evidence BRAF V600 mutations are uncommon in LGSOC; evidence is limited and largely derived from small or basket populations. ER/PR-positive ER/PR IHC Ribociclib + Letrozole Phase II Rationale is based on hormone-receptor expression and cyclin D-CDK4/6 biology, not recurrent CDK4/CDK6 mutations. EAOC ARID1A mutation or loss Tumor NGS panel and/or ARID1A IHC Tulmimetostat (EZH1/2i); Tuvusertib (ATRi) Phase I/II for EZH1/2i and phase I basket evidence for ATRi ARID1A alteration provides a potential vulnerability for EZH1/2 inhibition and ovarian evidence is focused mainly on ARID1A-mutated OCCC, with EnOC less well characterized; evidence for ATRi remains early-phase and basket-based; histotype-specific efficacy requires validation. MSI-H / dMMR MMR IHC and/or MSI testing; NGS where available; PD-L1 IHC Pembrolizumab; Nivolumab; Nivolumab + Ipilimumab; Sintilimab + Bevacizumab; Durvalumab Phase II MSI-H/dMMR is uncommon but clinically relevant; pembrolizumab has biomarker-defined tumor-agnostic approval, whereas ovarian-histotype-specific data remain limited. MOC ERBB2 amplification or HER2 overexpression HER2 IHC ± ISH and/or NGS copy-number profiling IBI354 Phase III ongoing HER2 expression is being tested in platinum-resistant ovarian, fallopian tube, or primary peritoneal cancer; MOC-specific efficacy remains unproven; accurate exclusion of metastatic gastrointestinal mucinous carcinoma is essential. Integrated histotype-molecular framework for therapeutic stratification in epithelial ovarian cancer Phase I; Phase II ongoing Berzosertib (ATRi); Adavosertib (WEE1i); Prexasertib (CHK1i) Adavosertib (WEE1i); ATRi + WEE1i Trametinib; Binimetinib; Selumetinib Tulmimetostat (EZH1/2i); Tuvusertib (ATRi) MMR IHC and/or MSI testing; NGS where available; PD-L1 IHC Pembrolizumab; Nivolumab; Nivolumab + Ipilimumab; Sintilimab + Bevacizumab; Durvalumab This framework illustrates both histotype-specific therapeutic opportunities and differences in evidentiary maturity. In HGSOC, BRCA1/2 mutation, HRD, and FRα expression have established clinical utility, whereas ATM-CHK2-p53 pathway deficiency, TP53 Y220C mutation and CCNE1 amplification remain investigational. In LGSOC, KRAS mutation is clinically actionable, while other MAPK pathway alterations, BRAF V600 mutation, and ER positivity represent additional therapeutic opportunities with varying levels of supporting evidence. In EAOC, ARID1A alterations and MSI-H/dMMR provide potential biomarker-defined therapeutic strategies, whereas ERBB2/HER2 alterations represent a potential therapeutic opportunity in MOC, although histotype-specific efficacy remains to be established. Thus, this framework is intended to provide a structured approach for interpreting molecular findings within histologic context, rather than treating all detected alterations as equivalent therapeutic targets.

A targeted literature search was performed using PubMed/MEDLINE and ClinicalTrials.gov. The search primarily covered literature published from 2006 through August 2026, with earlier seminal studies included where relevant. Search terms included combinations of “epithelial ovarian cancer,” “histotype,” “molecular subtype,” “targeted therapy,” “PARP,” “ATR,” “CHK1,” “WEE1,” “TP53,” “CCNE1,” “FOLR1,” “KRAS,” “BRAF,” “MEK,” “ARID1A,” “PIK3CA,” “EZH2,” “MMR,” “MSI,” “ERBB2,” and “HER2,” together with histotype-specific terms including “high-grade serous,” “low-grade serous,” “clear cell,” “endometrioid,” and “mucinous.” ClinicalTrials.gov was additionally searched using relevant drug names, molecular targets, and NCT identifiers. We prioritized peer-reviewed studies, pivotal or updated clinical trial reports, regulatory information, and translational studies that supported histotype-molecular therapeutic stratification. Trial status and available primary endpoint results were updated according to information available through August 2026.

Compared with other ovarian cancer histotypes, ovarian mucinous carcinoma (MOC) represents a relatively rare form of epithelial ovarian carcinoma and, owing to both its low incidence and the frequent pathological misclassification as gastrointestinal metastatic mucinous adenocarcinoma, the conduct of large-scale prospective clinical trials has been markedly constrained, thereby resulting in a scarcity of evidence-based data for MOC and causing its targeted therapeutic development to lag behind that of more prevalent histotypes. The GOG-0241 Phase III trial, which was designed to enroll patients with recurrent MOC, was terminated after accruing only 50 patients due to persistent enrollment challenges, and subsequent pathological review revealed that merely 45% of these cases were confirmed as primary MOC [ 20 ]. Pathologically, MOC is highly susceptible to confusion with metastatic mucinous adenocarcinoma of gastrointestinal origin, as demonstrated by retrospective analyses indicating that approximately two-thirds of cases initially diagnosed as primary MOC were reclassified as metastatic carcinoma upon expert pathological reassessment [ 156 ]. Genomic and immunohistochemical analyses collectively demonstrate that MOC exhibits a low incidence of homologous recombination deficiency, thereby suggesting a likely suboptimal response to PARPi. However, MOC frequently carries ERBB2 amplification (26.7%) and BRAF mutations (9%) [ 157 – 159 ] , indicating that MOC cannot adhere to the conventional treatment paradigm established for HGSOC but alternative molecularly informed targeted strategies instead. The ERBB2 gene encodes human epidermal growth factor receptor 2 (HER2), a member of the ERBB family of receptor tyrosine kinases which lacks an identified ligand yet forms stable heterodimers that constitutively activate downstream signaling pathways, including PI3K/AKT and RAS/MAPK pathways, thereby promoting tumor cell proliferation, survival, invasion, and metastasis [ 160 ]. In clinical practice HER2 has been regarded as a target of antibody-drug conjugate (Fig. 1 d). The HeriCare-Ovarian01 study ( NCT06834672 ) [ 161 ], which represents China’s first Phase III trial evaluating the efficacy of the antibody-drug conjugate (ADC) IBI354 in patients with HER2 expressing, platinum-resistant ovarian, primary peritoneal, or fallopian tube cancers, is currently enrolling. Patients with MOC may be included in subgroup analyses, which may ultimately provide clinically meaningful insights for future MOC-specific therapeutic strategies. Moreover, for MOC patients exhibiting ERBB2 amplification, the therapeutic rationale may extend to the use of established anti-HER2 agents already validated in breast and gastric cancers, such as trastuzumab, pertuzumab and ADC-based monotherapies or combination regimens [ 162 – 164 ].

The mitogen-activated protein kinases (MAPK) cascade is one of the most ancient and evolutionarily conserved signaling pathways, with three major MAPKs in mammals: the extracellular signal-regulated protein kinases (ERK), the p38 MAPK, and the c-Jun NH2-terminal kinases (JNK). The primary function of MAPK family is to transmit critical cell-proliferation signals from the cell surface to the nucleus [ 96 , 97 ]. Among these, the ERK MAPK pathway represents a central regulator of cell proliferation, activated through the RAS-RAF-MEK-ERK cascade, which conveys extracellular growth signals to promote cellular proliferation and differentiation [ 98 ]. A meta-analysis evaluating inhibitors of the RAS/RAF/MEK/ERK pathway in ovarian cancer demonstrated superior efficacy in LGSOC compared with other histotypes or unclassified ovarian cancers [ 99 ] , further supporting the relevance of integrating histotype with molecularly precision therapy. The RAS family encompasses KRAS, HRAS, and NRAS, and genomic analyses reveal frequent activating mutations in the RAS-RAF-MEK-ERK pathway in LGSOC, most commonly KRAS mutations (33.0%), NRAS (11.2%) and BRAF mutations (8.4%) [ 12 , 14 , 100 , 101 ]. KRAS exists in active and inactive conformations and is regulated as follows: growth factors stimulate receptor tyrosine kinases (RTKs), which subsequently activate KRAS proteins through intermediary molecules, thereby initiating downstream signaling cascades such as KRAS-BRAF-MEK-ERK or PI3K-AKT-mTOR [ 102 ]. Preclinical studies and trials confirm that inhibitors specifically targeting KRAS mutations reduce tumor cell viability, with ongoing investigation in non-small cell lung cancer (NSCLC) and pancreatic ductal adenocarcinoma (PDAC) [ 103 , 104 ]. In LGSOC, the more prevalent therapeutic strategy targeting the RAS-RAF-MEK-ERK pathway involves inhibition of KRAS downstream effectors (Fig. 1 b), particularly MEK. (Table 4 ) Table 4 Representative targeted therapy clinical trials in LGSOC Clinical Trial Study Identifier Agents Molecular Profile Target Setting Design Patients Primary Endpoint Results/States GOG-0239 [ 105 ] NCT00551070 Selumetinib RAS-RAF mutation status as review analysis MEK Recurrent disease Phase II, Single Group Assignment, Open Label Recurrent low grade serous carcinoma Tumor Response, Adverse Events ORR= 15.4% (8/52; 90% CI 7.9%-26.1%) DCO: Not reported MILO [ 106 , 107 ] NCT01849874 Binimetinib vs PCC KRAS mutation status as post hoc analysis MEK1/2 Recurrent disease Phase II, Randomized, Open Label LGS carcinoma of the ovary, fallopian tube or primary peritoneum PFS Primary analysis: PFS: 9.1 vs 10.6 mo, HR=1.21 (95% CI, 0.79-1.86) DCO: 20 Jan 2016 Trial stopped for futility Post hoc KRAS analysis: KRAS-mut vs KRAS-WT CR/PR with binimetinib, 44% vs 19% OR=3.4 (95% CI, 1.57-7.67) DCO: Aug 2021 GOG-0281 [ 108 ] NCT02101788 Trametinib vs standard-of-care MAPK pathway mutation status as post hoc analysis MEK1/2 Recurrent disease Phase II/III, Randomized, Open Label Recurrent low grade serous carcinoma PFS mPFS: 13.0 mo vs 7.2 mo HR=0.48 (95% CI, 0.36-0.64) DCO: 16 Jul 2019 RAMP 201 [ 109 ] NCT04625270 Avutometinib vs Avutometinib + Defactinib KRAS mutational status RAF/MEK FAK Recurrent disease Phase II, Randomized, Open Label Recurrent LGSOC (ovarian, peritoneal) ORR Combination regimen ORR: 31% (95% CI, 23-41); stratified KRAS subgroup ORR: KRASmut vs KRASwt ORR: 44% vs 17% DCO: 30 Jun 2024 RAMP 301 [ 110 ] NCT06072781 Avutometinib + Defactinib vs ICT KRAS mutational status RAF/MEK FAK Recurrent disease Phase III, Randomized, Open Label Recurrent LGSOC (ovarian, peritoneal) PFS Recruiting VE-BASKET [ 111 ] NCT01524978 Vemurafenib BRAF V600 mutation BRAF Resistant to normal therapy Phase II, Non-Randomized, Open Label Histologically confirmed solid cancers or multiple myeloma with a BRAF V600 mutation BORR Confirmed ORR 50.0% (95% CI 6.8%-93.2%); Ovarian cohort n=4; DCO: 12 Jan 2017 GOG-3026 [ 112 ] NCT03673124 Ribociclib + Letrozole - CDK4/6, ER/PR Recurrent disease Phase II, Single Group Assignment, Open Label Recurrent LGSOC Response rate ORR=30.6% (95% CI, 19.9-43.2) CBR=84% (95% CI, 72.5-91.6) DCO: 23 Dec 2024 MC1561 [ 113 ] NCT02657928 Ribociclib + Letrozole ER+, RB mutation status as post hoc analysis CDK4/6, ER/PR Recurrent disease Phase II, Single Group Assignment, Open Label Histologically confirmed recurrent ovarian, fallopian tube or primary peritoneal carcinoma or endometrial cancer in post-menopausal women PFS EC: mPFS= 5.4 mo OC: mPFS= 2.8 mo DCO: Not reported PFS, Progression-Free Survival; BORR, Best Overall Response Rate; ORR, Objective Response Rate; HR, Hazard Ratio; EC, Endometrial cancer; OC, Ovarian cancer; ICT, investigator's choice of treatments; CBR, clinical benefit rate; CI, confidence interval; DCO, data cut-off. Data cut-off dates are shown when reported in the original trial publications Representative targeted therapy clinical trials in LGSOC ORR= 15.4% (8/52; 90% CI 7.9%-26.1%) DCO: Not reported Primary analysis: PFS: 9.1 vs 10.6 mo, HR=1.21 (95% CI, 0.79-1.86) DCO: 20 Jan 2016 Trial stopped for futility Post hoc KRAS analysis: KRAS-mut vs KRAS-WT CR/PR with binimetinib, 44% vs 19% OR=3.4 (95% CI, 1.57-7.67) DCO: Aug 2021 mPFS: 13.0 mo vs 7.2 mo HR=0.48 (95% CI, 0.36-0.64) DCO: 16 Jul 2019 RAF/MEK FAK Combination regimen ORR: 31% (95% CI, 23-41); stratified KRAS subgroup ORR: KRASmut vs KRASwt ORR: 44% vs 17% DCO: 30 Jun 2024 RAF/MEK FAK Confirmed ORR 50.0% (95% CI 6.8%-93.2%); Ovarian cohort n=4; DCO: 12 Jan 2017 CDK4/6, ER/PR ORR=30.6% (95% CI, 19.9-43.2) CBR=84% (95% CI, 72.5-91.6) DCO: 23 Dec 2024 ER+, RB mutation status as post hoc analysis CDK4/6, ER/PR EC: mPFS= 5.4 mo OC: mPFS= 2.8 mo DCO: Not reported PFS, Progression-Free Survival; BORR, Best Overall Response Rate; ORR, Objective Response Rate; HR, Hazard Ratio; EC, Endometrial cancer; OC, Ovarian cancer; ICT, investigator's choice of treatments; CBR, clinical benefit rate; CI, confidence interval; DCO, data cut-off. Data cut-off dates are shown when reported in the original trial publications Clinical trials investigating the efficacy of MEK inhibitor (MEKi) in LGSOC across different mutational contexts have been ongoing since 2013. The GOG-0239 single-arm trial enrolled 52 patients with recurrent LGSOC to evaluate treatment response to the MEKi selumetinib, demonstrating complete or partial responses in 15.4% of patients. However, subsequent analysis of RAS-RAF mutations revealed no differential response rates across mutational subgroups, indicating that further exploration of predictive biomarkers was required [ 105 ]. Subsequently, randomized controlled trials (RCTs) for LGSOC were initiated, including the MILO study, which compared binimetinib, a potent MEK1/2 inhibitor, to physician’s choice chemotherapy (PCC). The trial was terminated early after enrolling 341 patients due to predefined futility boundaries, though post hoc analyses suggested a potential association between KRAS mutation status and binimetinib response, with KRAS-mutant patients exhibiting significantly higher objective response rates than KRAS-wild-type patients (44% vs. 19%) [ 106 , 107 ], reinforcing the investigational value of MEKi in LGSOC. The GOG-0281 phase II/III RCT compared trametinib to standard therapy in recurrent LGSOC, demonstrating superior efficacy for trametinib, although no correlation with KRAS, BRAF, or NRAS mutation status was observed, possibly due to MAPK pathway activation, rather than classic mutations alone [ 108 ]. The RAMP201 trial evaluated the combination of the RAF/MEK inhibitor avutometinib and the FAK inhibitor defactinib versus avutometinib monotherapy in recurrent LGSOC. After establishing safety and preliminary efficacy, subgroup analysis by KRAS status revealed an overall ORR of 31%, with KRAS-mutant patients achieving substantially higher ORR compared with KRAS wild-type patients (44% vs. 17%) [ 109 , 114 ]. On the basis of the RAMP201 data, the FDA granted accelerated approval in May 2025 to avutometinib plus defactinib for adult patients with KRAS-mutated recurrent LGSOC patients who had received prior systemic therapy. This approval supports KRAS mutation as a clinically actionable stratification marker for this combination in recurrent LGSOC. The therapeutic role of this combination is being further evaluated in the ongoing randomized phase III RAMP301 trial, which specifically enrolls patients with recurrent LGSOC and compares avutometinib plus defactinib with investigator’s choice of treatment. The results of RAMP301 trial may provide further prospective evidence for the role of RAF/MEK-FAK inhibition in LGSOC [ 110 ]. In addition to MEK inhibitor (MEKi), several studies have also examined the therapeutic potential of BRAF inhibitor (BRAFi), as the most common BRAF alteration, the V600E point mutation, disrupts BRAF autoinhibition and thereby induces sustained activation of downstream signaling [ 115 ], rendering BRAFi another targeted strategy against the RAS/RAF/MEK/ERK cascade alongside MEKi. However, clinical trials specifically evaluating BRAFi in LGSOC remain limited. The VE-BASKET trial recruited patients carrying BRAF V600 mutations to assess the efficacy of vemurafenib across solid tumors, and the results demonstrated preliminary antitumor activity of vemurafenib in BRAF V600-mutant non-small cell lung cancer, Erdheim-Chester disease, and Langerhans cell histiocytosis, thereby underscoring its potential as a molecularly targeted therapy. One LGSOC patient achieved a durable partial response (12.9 mo), although the small sample size needs cautious interpretation [ 111 ]. Beyond monotherapy, LGSOC may also benefit from BRAFi-MEKi combination regimens, which has already been validated in refractory BRAF-mutant solid tumors, as evidenced by an ongoing clinical trial ( NCT05355701 ) [ 116 ]. In addition to regulation by cell cycle checkpoints, cell cycle is also tightly correlated to cyclin-dependent kinase (CDK) activity, as exogenous mitogenic signals promote cyclin D1 (CCND1) binding and activating CDK4/6, which subsequently phosphorylate transcriptional repressors such as p130 and p107, thereby releasing transcription factors including E2F and reinforcing the positive feedback loops that collectively drive the G1-S phase transition [ 117 , 118 ] (Fig. 3 ). Genomic analyses further reveal frequent CCND1 amplification across multiple human cancers [ 119 ], with CCND1, CCND2, CCND3, CDK4, and CDK6 predominantly amplified in ovarian cancer [ 118 ] , suggesting the therapeutic rationale for CDK4/6 inhibition in this malignancy. Moreover, most LGSOC tissues express estrogen receptor (ER) and progesterone receptor (PR) [ 120 , 121 ] , which has been proved to be correlated to CCND1 expression level [ 122 , 123 ]. Clinical trials have demonstrated the efficacy of hormonal maintenance therapy (HMT) in LGSOC patients [ 124 ], thereby providing the biological and clinical basis for integrating CDK4/6 inhibitors with endocrine therapy as a common targeted strategy in LGSOC-related clinical trials. Fig. 3 Mechanism of combined CDK4/6 inhibition and endocrine therapy in LGSOC. In ER/PR-positive LGSOC cells, estradiol (E2) binding activates the estrogen receptor (ER), leading to transcriptional upregulation of CCND1 (encoding Cyclin D1). Cyclin D1 binds to and activates CDK4/6, which phosphorylates pocket proteins (Rb, p130 and p107), leading to E2F release, downstream CCNE expression, and G1/S checkpoint progression. Created with BioGDP.com [ 21 ] Mechanism of combined CDK4/6 inhibition and endocrine therapy in LGSOC. In ER/PR-positive LGSOC cells, estradiol (E2) binding activates the estrogen receptor (ER), leading to transcriptional upregulation of CCND1 (encoding Cyclin D1). Cyclin D1 binds to and activates CDK4/6, which phosphorylates pocket proteins (Rb, p130 and p107), leading to E2F release, downstream CCNE expression, and G1/S checkpoint progression. Created with BioGDP.com [ 21 ] The NCT02657928 trial collected 20 endometrial cancer and 20 ovarian cancer patients to evaluate CDK4/6 inhibitor ribociclib plus letrozole, and within the ovarian cancer subgroup, comprising 3 LGSOC and 17 HGSOC cases, four patients (20%) achieved 24-week PFS, of which three responders were LGSOC (one complete response and two partial responses lasting more than two years), whereas only one of the 17 HGSOC patients remained progression-free for ≥24 weeks, thereby indicating preferential therapeutic benefit in LGSOC, rather than HGSOC [ 113 ]. Furthermore, the ongoing GOG-3026 trial evaluating ribociclib combined with endocrine therapy in recurrent LGSOC has reported outcomes showing an ORR of 30.6% and a clinical benefit rate of 84% [ 112 ].

Retrospective studies have demonstrated that substantial differences exist in the biological and clinical characteristics of different histotypes of ovarian cancer, yet standardized treatment continues to be surgery combined with platinum-based chemotherapy, thereby resulting in remarkable heterogeneity in clinical response rates among these subtypes. This review synthesizes the molecular targets in current clinical practice and under investigation for epithelial ovarian cancer on the basis of histological classification, with the aim of integrating histological and molecular profiling to inform the development of novel, precision-guided therapeutic strategies tailored to this highly heterogeneous disease. To this end, we propose a unified histotype-molecular framework that links histologic classification and molecular characterization with biomarker assessment, therapeutic selection, and the level of supporting clinical evidence. Although this framework provides a structured approach for linking histotype and molecular features with potential therapeutic strategies, its clinical value depends on whether these elements can be reliably implemented in routine practice. Several practical barriers should therefore be considered. First, accurate histotype assignment is a prerequisite for appropriate biomarker testing and treatment selection. Although integrated morphology and IHC-based algorithms can improve ovarian carcinoma histotyping, misclassification remains possible, particularly for uncommon histotypes such as LGSOC and MOC [ 165 , 166 ]. This issue is clinically relevant in MOC, where central pathological review of advanced-stage cases showed that only approximately one third of tumors initially classified as primary mucinous ovarian carcinoma were interpreted as true primary ovarian tumors [ 156 ]. Second, biomarker assays differ substantially in feasibility and clinical logistics. IHC-based tests, including FRα, MMR proteins, ER/PR, and HER2, can be incorporated into routine pathology workflows using FFPE sections, whereas NGS panels and HRD assays require extraction of adequate-quality nucleic acids from tumor-rich specimens and are more dependent on tumor cellularity, sample quantity, fixation quality, cost, and turnaround time. Published implementation studies indicate that HRD testing turnaround times vary considerably across settings, ranging from approximately one week for selected in-house assays to several weeks in centralized or routine clinical workflows, and the associated testing costs may not always be covered by medical insurance [ 167 – 169 ]. Third, the evidence supporting matched therapies is heterogeneous, with many non-approved strategies relying on single-arm trials, basket studies, post hoc subgroup analyses, or small histotype-specific cohorts. Therefore, implementation of this framework requires integration of diagnostic reliability, assay feasibility, tissue adequacy, and evidence maturity. To address these limitations, several strategies may improve the clinical feasibility of histotype-molecular stratification. For diagnostic implementation, expert gynecologic pathology review and selected IHC panels may reduce histotype misclassification, particularly in diagnostically challenging cases such as MOC. Commonly used markers include WT1 and p53 for serous carcinoma, napsin A and HNF1B for clear cell carcinoma, ER/PR for endometrioid or hormone receptor assessment, and CK7/CK20/CDX2/SATB2-based panels when distinguishing primary MOC from metastatic gastrointestinal tumors [ 165 , 166 ]. For molecular testing, standardized pre-analytic assessment, pathologist-guided or image-guided selection of representative tumor-rich areas, macrodissection when needed, and extraction from samples with sufficient tumor cellularity and nucleic-acid quality may improve NGS and HRD testing success [ 170 , 171 ]. In addition, tumor-only sequencing may identify BRCA1/2 or other homologous recombination repair gene alterations, but it cannot always determine whether a variant is somatic or germline in origin. Confirmation of germline status therefore requires paired analysis with non-tumor DNA or separate clinical germline testing, usually accompanied by genetic counseling, because germline findings have implications for hereditary cancer risk assessment and cascade testing of relatives [ 172 , 173 ]. In settings where centralized HRD testing is delayed, costly, or unavailable, reflex testing pathways, parallel germline/tumor testing, sequential testing adapted to local resources, and validated in-house HRD assays may help improve accessibility and turnaround time [ 172 – 174 ]. Finally, molecular tumor boards and evidence-ranking systems such as ESCAT may help prioritize alterations with established clinical actionability over investigational or preclinical targets, thereby reducing the risk of overinterpreting molecular findings with limited clinical actionability [ 175 ]. Taken together, the proposed histotype-molecular stratification framework should be regarded as an evolving model rather than a fixed clinical algorithm. Although its current application should be interpreted in light of differences in evidence maturity, testing feasibility, and the extent of prospective validation, these considerations also define key directions for future research. Prospective studies should evaluate whether stratification based on integrated histologic and molecular features can improve treatment selection and clinical outcomes. In parallel, translational studies are needed to refine biomarker combinations, clarify resistance mechanisms, and identify context-dependent therapeutic vulnerabilities across histotypes. For rare ovarian cancer histotypes, multicenter collaboration, histotype-specific expansion cohorts, and adaptive trial designs may be particularly important to generate clinically interpretable evidence. Overall, this review supports moving beyond a single treatment paradigm toward composite stratification models that integrate histology, molecular profiling, therapeutic actionability, and evidence maturity in EOC.

TP53 encodes the p53 protein that serves as a central regulator of genomic integrity, cellular stress responses, and cell-cycle control through regulating approximately 500 downstream genes [ 28 ]. TP53 mutations are among the most prevalent genetic alterations in HGSOC and are widespread across diverse tumor types. Despite remarkable frequency and biological importance, TP53 mutations were long considered pharmacologically inaccessible, reinforcing the perception of p53 as an “undruggable” target [ 29 ]. This view has been overturned as accumulating evidence demonstrates its therapeutic feasibility, including restoration of wild-type p53 activity, degradation of mutant p53, inhibition or inactivation of binding partners that potentiate mutant p53 function, blockade of mutant p53 downstream signaling pathways, exploitation of synthetic lethality, and mutant p53-specific immunotherapy [ 30 ]. Nevertheless, only two of these approaches have advanced into clinical trials. One of them aims to restore mutant p53 activity. Trial NCT03072043 tested the p53-reactivating molecule eprenetapopt combined with azacitidine in TP53-mutant hematologic malignancies and reported early clinical activity in a phase Ib/II setting [ 31 ]. However, the subsequent randomized phase III trial ( NCT03745716 ) in TP53-mutant myelodysplastic syndrome failed to meet its primary endpoint [ 32 , 33 ], indicating that eprenetapopt should be regarded as proof-of-concept rather than established clinical evidence. More recently, rezatapopt (PC14586), an oral selective p53 Y220C reactivator, showed preliminary antitumor activity in TP53 Y220C-mutated advanced solid tumors, including ovarian cancer ( NCT04585750 ) [ 34 ], and a phase II PYNNACLE interim result about ovarian cancer analysis further reported encouraging responses [ 35 ]. Nevertheless, rezatapopt remains investigational and is applicable only to the TP53 Y220C-mutated subset. The other strategy associated with TP53 mutation applies synthetic lethality to the ATM-CHK2-p53 pathway, which has broader relevance in ovarian cancer and will be detailed in the following part of cell-cycle checkpoints. Cells exposed to endogenous or exogenous insults sustain various types of DNA damage, including impaired base pairing, disruptions in DNA replication or transcription, base deletion, single-strand breaks (SSBs), and double-strand breaks (DSBs). Normal cells activate DNA damage repair (DDR) mechanisms to restore genomic integrity and maintain stability for transmission to daughter cells. These mechanisms include base excision repair and nucleotide excision repair for SSBs, mismatch repair, as well as non-homologous end-joining (NHEJ) and homologous recombination (HR) for DSBs [ 36 , 37 ]. Genomic instability is a hallmark of malignancy [ 38 – 40 ], and a defining feature of HGSOC. During the process of HGSOC progression, rapid DNA replication and cell division increase the possibility of replication-associated DNA damage. In DDR-deficient tumor cells, accumulating DNA damage during replication leads to a higher mutational burden and further genomic instability. The accumulation of pro-tumorigenic mutations or the loss of tumor-suppressive mutations enables malignant cells to evade immune surveillance, bypass senescence, and resist apoptosis, thereby promoting tumor progression. Conversely, DDR deficiency also creates therapeutic vulnerabilities in HGSOC. Dysfunction or mutation in one DDR pathway increases tumor cell dependence on alternative DDR pathways. Targeted inhibition of these compensatory pathways forces tumor cells with excessive DNA damage to enter mitosis prematurely, triggering cell death. Notably, normal cells without DDR deficiency tolerate such treatment without cytotoxicity, which is the synthetic lethality strategy originally proposed by geneticists [ 41 ]. PARP inhibitors are a canonical model of synthetic lethality, with their therapeutic rationale rooted in BRCA mutation [ 42 , 43 ], most prominently observed in HGSOC. BRCA1/2 are essential mediators of high-fidelity homologous recombination repair [ 44 ]. Approximately 21% of HGSOC cases harbor germline or somatic BRCA1/2 mutations [ 8 ], resulting in homologous recombination deficiency (HRD). PARP1, a central member of the PARP family, mediates single-strand break (SSB) repair and contributes to DSB repair through NHEJ and microhomology-mediated end-joining (MMEJ), while also stabilizing stalled replication forks [ 45 ]. On the occasion of DNA damage, PARP1 is rapidly recruited to the DNA damage lesion site and catalyzes poly (ADP-ribose) (pADPr) synthesis using NAD⁺ and ATP. The resulting pADPr chains form a platform that facilitates the recruitment of downstream DDR components, including HR-related proteins [ 37 ]. Hence, pharmacologic inhibition of PARP1 in HGSOC harboring intrinsic BRCA mutations precipitates replication-fork collapse and mitotic entry with unresolved lesions, leading to synthetic lethality ultimately (Fig. 2 a). Since the U.S. FDA approved the first PARP inhibitor (PARPi), olaparib, for deleterious or suspected deleterious germline BRCA-mutated advanced ovarian cancer after three or more prior lines of chemotherapy in 2014 [ 44 ], the use of PARPis has expanded rapidly, particularly in BRCA-mutated HGSOC [ 46 ]. Approved PARP inhibitors include FDA-approved olaparib, niraparib and rucaparib, as well as fluzoparib, pamiparib, and senaparib approved by China’s National Medical Products Administration (NMPA). Phase II trials such as ARIEL2 and QUADRA stratified HGSOC patients by BRCA1/2 mutation or HR status and reported objective response rates (ORR) of 23%-27% [ 47 , 48 ], establishing BRCA or HRD status as predictive biomarkers of PARPi response. These findings promoted subsequent Phase III trial designs, which demonstrated clinically meaningful benefits. Table 1 summarizes pivotal Phase III studies evaluating PARPi monotherapy or PARPi-based combinations in HGSOC. Among these trials, the ARIEL4 trial met its primary endpoint by demonstrating an improvement in PFS, though overall survival (OS) was in fact inferior to chemotherapy, suggesting that the use of rucaparib in recurrent or later-line BRCA-mutated ovarian cancer may not always represent the optimal therapeutic strategy [ 59 , 60 ]. Fig. 2 Targeted therapeutic strategies related to DNA damage repair and cell-cycle checkpoint vulnerabilities in HGSOC. The schematic illustrates key mechanisms: a) Targeting DNA damage repair with PARP inhibitors (PARPi) and BRD4 inhibitors (BRD4i), which can induce synthetic lethality. b) Reactivation of mutant p53. c) Targeting the G2/M checkpoint vulnerability caused by G1/S checkpoint abnormality using ATR/CHK1/WEE1 inhibitors. Created with BioGDP.com [ 21 ] Table 1 Key phase III clinical trials of PARP inhibitors in HGSOC Clinical Trial Study Identifier Agents Molecular Profile Target Setting Design Patients Primary Endpoint Results/States SOLO1 [ 49 , 50 ] NCT01844986 Olaparib vs Placebo BRCA1/2 mutated PARP1/2 Maintenance Phase III, Randomized, Quadruple-Blind HGSOC/EnOC/fallopian tube/peritoneal carcinoma PFS PFS: 56.0 vs 13.8 mo, HR=0.33 (95% CI, 0.25-0.43) DCO: 5 Mar 2020 SOLO2 [ 51 , 52 ] NCT01874353 Olaparib vs Placebo BRCA1/2 mutated PARP1/2 Platinum-sensitive recurrent Phase III, Randomized, Quadruple-Blind HGSOC, high grade EnOC PFS PFS: 19.1 vs 5.5 mo, HR=0.30 (95% CI, 0.22-0.41) DCO: 19 Sep 2016; OS: 51.7 vs 38.8 mo, HR=0.74 (95% CI, 0.54-1.00) DCO: 3 Feb 2020 PRIMA [ 53 ] NCT02655016 Niraparib vs Placebo HRD, HRP PARP1/2 Maintenance Phase III, Randomized, Quadruple-Blind HGSOC, high grade EnOC PFS HRD+ PFS: 21.9 vs 10.4 mo, HR=0.43 (95% CI, 0.31-0.59) DCO: 17 May 2019 PAOLA-1 [ 54 ] NCT02477644 Olaparib + Bevacizumab vs Placebo + Bevacizumab HRD, HRP PARP1/2, VEGF-A Maintenance Phase III, Randomized, Triple-Blind HGSOC, EnOC, other epithelial non mucinous ovarian cancer PFS HRD+ PFS: 37.2 vs 17.7 mo, HR=0.33 (95% CI, 0.25-0.45); ITT PFS: 22.1 vs 16.6 mo, HR=0.59 (95% CI, 0.49-0.72) DCO: 22 May 2019 NOVA [ 55 , 56 ] NCT01847274 Niraparib vs Placebo gBRCAmut, non-gBRCAmut PARP1/2 Platinum-sensitive recurrent Phase III, Randomized, Quadruple-Blind HGSOC or other with gBRCAmut PFS gBRCAmut PFS: 21.0 vs 5.5 mo, HR=0.27 (95% CI, 0.17-0.41); non-gBRCAmut PFS: 9.3 vs 3.9 mo, HR=0.45 (95% CI, 0.34-0.61) DCO: 20 Jun 2016 NORA [ 57 ] NCT03705156 Niraparib vs Placebo BRCA status as post hoc analysis PARP1/2 Platinum-sensitive recurrent Phase III, Randomized, Quadruple-Blind HGSOC PFS mPFS: 18.3 mo vs 5.4 mo HR=0.32 (95% CI, 0.23-0.45) DCO: 1 Feb 2020 ARIEL3 [ 58 ] NCT01968213 Rucaparib vs Placebo BRCA mutated, HRD, HRP PARP1/2 Platinum-sensitive recurrent Phase III, Randomized, Double-Blind HGSOC, EnOC, fallopian tube, or primary peritoneal cancer PFS BRCAm PFS: 16.6 vs 5.4 mo, HR=0.23 (95% CI, 0.16-0.34); HRD+ PFS: 13.6 vs 5.4 mo, HR=0.32 (95% CI, 0.24-0.42); ITT PFS: 10.8 vs 5.4 mo, HR=0.36 (95% CI, 0.30-0.45) DCO: 15 Apr 2017 ARIEL4 [ 59 , 60 ] NCT02855944 Rucaparib vs Chemotherapy BRCA1/2 mutation PARP1/2 Recurrent disease Phase III, Randomized, Open Label HGSOC, Grade 2/3 EnOC, fallopian tube, or primary peritoneal cancer PFS mPFS:7.4 mo vs 5.3 mo, HR=0.64 (95% CI, 0.49-0.84) DCO: 30 Sep 2020 OS:19.4 mo VS 25.4 mo HR=1.31 (95% CI, 1.00-1.70) DCO: 10 Apr 2022 ATHENA-MONO [ 61 ] NCT03522246 Rucaparib vs Placebo HRD, HRP PARP1/2 Maintenance Phase III, Randomized, Double-Blind Advanced EOC, fallopian tube, or primary peritoneal cancer PFS HRD+ PFS: 28.7 vs 11.3 mo, HR=0.47 (95% CI, 0.31-0.72); ITT PFS: 20.2 vs 9.2 mo, HR=0.52 (95% CI, 0.40-0.68) DCO: 23 Mar 2022 FLAMES [ 62 ] NCT04169997 Senaparib vs Placebo BRCAmut, BRCAwt PARP1/2 Maintenance Phase III, Randomized, Quadruple-Blind HGSOC or other ovarian cancer, fallopian tube cancer or primary peritoneal cancer with BRCA mutation PFS BRCAmut PFS: NR vs 15.6 mo, HR=0.43 (95% CI, 0.24-0.76) ITT PFS: NR vs 13.6 mo, HR=0.43 (95% CI, 0.32-0.58) DCO: 16 Mar 2023 VELIA [ 63 ] NCT02470585 Veliparib vs Placebo BRCA status/ HRD, HRP PARP1/2 First-line + Maintenance Phase III, Randomized, Double-Blind Advanced EOC PFS Veliparib-throughout versus control: HRD vs BRCAwt PFS: 22.9 mo vs 19.8 mo, HR=0.76 (95% CI, 0.53-1.09) DCO: 3 May 2019 HRP, homologous recombination proficient; PFS, Progression-Free Survival; HR, Hazard Ratio; ITT, Intention-To-Treat; CI, confidence interval; DCO, data cut-off. Data cut-off dates are shown when reported in the original trial publications Targeted therapeutic strategies related to DNA damage repair and cell-cycle checkpoint vulnerabilities in HGSOC. The schematic illustrates key mechanisms: a) Targeting DNA damage repair with PARP inhibitors (PARPi) and BRD4 inhibitors (BRD4i), which can induce synthetic lethality. b) Reactivation of mutant p53. c) Targeting the G2/M checkpoint vulnerability caused by G1/S checkpoint abnormality using ATR/CHK1/WEE1 inhibitors. Created with BioGDP.com [ 21 ] Key phase III clinical trials of PARP inhibitors in HGSOC Phase III, Randomized, Quadruple-Blind PFS: 56.0 vs 13.8 mo, HR=0.33 (95% CI, 0.25-0.43) DCO: 5 Mar 2020 PFS: 19.1 vs 5.5 mo, HR=0.30 (95% CI, 0.22-0.41) DCO: 19 Sep 2016; OS: 51.7 vs 38.8 mo, HR=0.74 (95% CI, 0.54-1.00) DCO: 3 Feb 2020 HRD+ PFS: 21.9 vs 10.4 mo, HR=0.43 (95% CI, 0.31-0.59) DCO: 17 May 2019 Olaparib + Bevacizumab vs Placebo + Bevacizumab PARP1/2, VEGF-A HRD+ PFS: 37.2 vs 17.7 mo, HR=0.33 (95% CI, 0.25-0.45); ITT PFS: 22.1 vs 16.6 mo, HR=0.59 (95% CI, 0.49-0.72) DCO: 22 May 2019 gBRCAmut, non-gBRCAmut gBRCAmut PFS: 21.0 vs 5.5 mo, HR=0.27 (95% CI, 0.17-0.41); non-gBRCAmut PFS: 9.3 vs 3.9 mo, HR=0.45 (95% CI, 0.34-0.61) DCO: 20 Jun 2016 mPFS: 18.3 mo vs 5.4 mo HR=0.32 (95% CI, 0.23-0.45) DCO: 1 Feb 2020 BRCAm PFS: 16.6 vs 5.4 mo, HR=0.23 (95% CI, 0.16-0.34); HRD+ PFS: 13.6 vs 5.4 mo, HR=0.32 (95% CI, 0.24-0.42); ITT PFS: 10.8 vs 5.4 mo, HR=0.36 (95% CI, 0.30-0.45) DCO: 15 Apr 2017 mPFS:7.4 mo vs 5.3 mo, HR=0.64 (95% CI, 0.49-0.84) DCO: 30 Sep 2020 OS:19.4 mo VS 25.4 mo HR=1.31 (95% CI, 1.00-1.70) DCO: 10 Apr 2022 HRD+ PFS: 28.7 vs 11.3 mo, HR=0.47 (95% CI, 0.31-0.72); ITT PFS: 20.2 vs 9.2 mo, HR=0.52 (95% CI, 0.40-0.68) DCO: 23 Mar 2022 BRCAmut, BRCAwt BRCAmut PFS: NR vs 15.6 mo, HR=0.43 (95% CI, 0.24-0.76) ITT PFS: NR vs 13.6 mo, HR=0.43 (95% CI, 0.32-0.58) DCO: 16 Mar 2023 Veliparib vs Placebo Veliparib-throughout versus control: HRD vs BRCAwt PFS: 22.9 mo vs 19.8 mo, HR=0.76 (95% CI, 0.53-1.09) DCO: 3 May 2019 HRP, homologous recombination proficient; PFS, Progression-Free Survival; HR, Hazard Ratio; ITT, Intention-To-Treat; CI, confidence interval; DCO, data cut-off. Data cut-off dates are shown when reported in the original trial publications Despite the established benefit of PARPi in BRCA-mutated or HRD-positive HGSOC, both primary and acquired resistance remain common clinical challenges. Mechanistically, PARPi resistance can be broadly attributed to PARP-associated alterations, restoration of HRR function, and recovery of replication fork stability. PARP-associated resistance may arise through PARG deficiency, which restores PARylation and reduces PARPi-induced PARP1-DNA complex formation, whereas HRR restoration can occur through secondary BRCA1/2 reversion mutations or loss of DNA end-protection factors such as the Shieldin complex. Fork-stabilizing pathways, including MSANTD4-related mechanisms, may further bypass BRCA1/2-RAD51 dependency. The ATROVA trial is evaluating ctDNA-based monitoring of BRCA1/2 functional restoration as a strategy for early detection of PARPi resistance. Relevant PARPi-based combination strategies currently under clinical investigation are summarized in Table 2 . Table 2 Representative Clinical Trials of PARP Inhibitors Combination Strategies in Ovarian Cancer Clinical Trial Study Identifier Agents Molecular Profile Target Setting Design Patients Primary Endpoint Results/States 12-159 [ 64 ] NCT01623349 Olaparib + Alpelisib gBRCAm, Non gBRCAm PARP1/2, PI3K Recurrent disease Phase Ib, Single Group Assignment, Open Label HGSOC or gBRCAm EOC, triple negative BC or other gBRCAm tumor RP2D, MTD ORR=35.7% (10/28; all partial responses) DCO: Not reported EPIK-O [ 65 ] NCT04729387 Olaparib + Alpelisib vs TPC Non gBRCAm PARP1/2, PI3K Recurrent disease Phase III, Randomized, Open Label HGSOC, high-grade EnOC, OCCC PFS The primary endpoint was not met DCO: 21 Apr 2023 16-573 [ 66 ] NCT03057145 Olaparib + Prexasertib BRCA1/2 mut PARP1/2, CHK1 Recurrent disease Phase I, Single Group Assignment, Open Label PARP inhibitor-resistant HGSOC MTD BRCA mut: 22% PR DCO: Not reported MEDIOLA [ 67 ] NCT02734004 Olaparib + Durvalumab ± Bevacizumab gBRCAm, Non gBRCAm PARP1/2, PD-L1 VEGF-A Platinum-sensitive recurrent Phase II, Single Group Assignment, Open Label Small cell lung cancer, HER2-negative breast cancer, ovarian cancer, gastric cancer ORR, DCR at 24w gBRCAm doublet: ORR= 92.2% (95% CI 81.1-97.8) non-gBRCAm doublet: DCR= 28.1% (90% CI 15.5-43.9) non-gBRCAm triplet: DCR= 74.2% (90% CI 58.2-86.5) DCO: 17 Sep 2021 NRG-GY005 [ 68 ] NCT02502266 Olaparib + Cediranib vs Cediranib vs Olaparib vs Chemotherapy BRCA PARP1/2, VEGF-A Platinum-resistant or primary platinum-refractory disease Phase II/III, Randomized, Open Label HGSOC, high grade EnOC PFS Cediranib/Olaparib: PFS=5.2 mo Cediranib: PFS=4.0 mo SOC: PFS=3.4 mo DCO: 13 Jun 2023 DUO-O [ 69 ] NCT03737643 First-line : CTx + Bevacizumab ± Durvalumab Maintenance : Bevacizumab ± Durvalumab ± Olaparib tBRCAm status VEGF-A PD-L1 PARP1/2 First-line/ Maintenance Phase III, Randomized, Quadruple-blind HGSOC, EnOC, OCCC, Carcinosarcoma PFS non-tBRCAm HRD positive PFS: 45.1 mo vs 23.3 mo, HR=0.46 (95% CI, 0.33-0.65); non-tBRCAm HRD positive ITT PFS: 25.1 mo vs 19.3 mo, HR=0.61 (95% CI, 0.51-0.73) DCO: 18 Sep 2023 KEYLYNK-001 [ 70 ] NCT03740165 Pembrolizumab + Olaparib vs Pembrolizumab + Placebo vs Placebo + Placebo BRCA1/2-nonmutated PD-1, PARP1/2 First-line / Maintenance Phase III, Randomized, Quadruple-blind Stage III or Stage IV epithelial ovarian cancer PFS PFS: 22.2 vs 14.6 mo; HR=0.71 (95% CI, 0.61-0.84) DCO: Not reported FIRST [ 71 ] NCT03602859 Niraparib ± Dostarlimab ± Bevacizumab HRR status PD-L1 status PD-1, PARP1/2 First-line/ Maintenance Phase III, Randomized, Triple -blind High-grade nonmucinous epithelial ovarian cancer PFS PFS: 20.63 vs 19.19 mo; HR=0.85 (95% CI, 0.73-0.99) DCO: 31 Oct 2024 PFS, Progression-Free Survival; PR, Partial Response; ORR, Objective Response Rate; DCR, Disease Control Rate; SOC, standard-of-care chemotherapy; ITT, Intention-To-Treat; HR, Hazard Ratio; CI, confidence interval; DCO, data cut-off. Data cut-off dates are shown when reported in the original trial publications Representative Clinical Trials of PARP Inhibitors Combination Strategies in Ovarian Cancer gBRCAm, Non gBRCAm PARP1/2, PI3K RP2D, MTD ORR=35.7% (10/28; all partial responses) DCO: Not reported Olaparib + Alpelisib vs TPC PARP1/2, PI3K The primary endpoint was not met DCO: 21 Apr 2023 PARP1/2, CHK1 BRCA mut: 22% PR DCO: Not reported gBRCAm, Non gBRCAm PARP1/2, PD-L1 VEGF-A ORR, DCR at 24w gBRCAm doublet: ORR= 92.2% (95% CI 81.1-97.8) non-gBRCAm doublet: DCR= 28.1% (90% CI 15.5-43.9) non-gBRCAm triplet: DCR= 74.2% (90% CI 58.2-86.5) DCO: 17 Sep 2021 Olaparib + Cediranib vs Cediranib vs Olaparib vs Chemotherapy PARP1/2, VEGF-A Cediranib/Olaparib: PFS=5.2 mo Cediranib: PFS=4.0 mo SOC: PFS=3.4 mo DCO: 13 Jun 2023 First-line : CTx + Bevacizumab ± Durvalumab Maintenance : Bevacizumab ± Durvalumab ± Olaparib VEGF-A PD-L1 PARP1/2 HGSOC, EnOC, OCCC, Carcinosarcoma non-tBRCAm HRD positive PFS: 45.1 mo vs 23.3 mo, HR=0.46 (95% CI, 0.33-0.65); non-tBRCAm HRD positive ITT PFS: 25.1 mo vs 19.3 mo, HR=0.61 (95% CI, 0.51-0.73) DCO: 18 Sep 2023 Pembrolizumab + Olaparib vs Pembrolizumab + Placebo vs Placebo + Placebo PD-1, PARP1/2 PFS: 22.2 vs 14.6 mo; HR=0.71 (95% CI, 0.61-0.84) DCO: Not reported Niraparib ± Dostarlimab ± Bevacizumab HRR status PD-L1 status PD-1, PARP1/2 Phase III, Randomized, Triple -blind PFS: 20.63 vs 19.19 mo; HR=0.85 (95% CI, 0.73-0.99) DCO: 31 Oct 2024 PFS, Progression-Free Survival; PR, Partial Response; ORR, Objective Response Rate; DCR, Disease Control Rate; SOC, standard-of-care chemotherapy; ITT, Intention-To-Treat; HR, Hazard Ratio; CI, confidence interval; DCO, data cut-off. Data cut-off dates are shown when reported in the original trial publications BRD4, a member of the bromodomain and extra-terminal (BET) protein family, is characterized by two tandem bromodomains (BDs) and an extra-terminal domain (ETD). BRD4 exists in two isoforms, BRD4-L and BRD4-S, and functions as an epigenetic biomarker by binding acetylated core histones and regulating gene expression [ 72 ]. Additionally, BRD4 plays a role in DDR process. When DSB event occurs, increased acetylation of histone H4 and phosphorylation of H2AX facilitate the recruitment of BRD4-L to damaged DNA sites. BRD4-L serves as an anchoring platform for DNA repair proteins, promoting the assembly of DNA repair and activating the NHEJ pathway [ 73 ] (Fig. 2 a). Consequently, BRD4 inhibition leads to impaired DNA damage repair. Preclinical studies have demonstrated that BRD4 inhibitors induce homologous recombination deficiency, which can be used as a synthetic lethal strategy alongside PARP inhibitors [ 74 ]. Clinical trials exploring BRD4 inhibition are gradually emerging, though most remain in early-phase or basket trial stages. A basket trial ( NCT03205176 ) included 14 platinum-resistant HGSOC/fallopian tube/primary peritoneal cancer patients, aiming to evaluate the efficacy of the BRD4 inhibitor AZD5153 combined with the PARPi olaparib in patients with relapsed/refractory solid tumors or lymphomas [ 75 ]. The efficacy in HGSOC cases warrants further investigation, though one patient with metastatic BRCA-negative pancreatic ductal adenocarcinoma of this trial achieved a partial response lasting 4.2 months, suggesting potential clinical benefit. (Table 3 ) Table 3 Representative Non-PARPi Targeted Therapy Clinical Trials in HGSOC Clinical Trial Study Identifier Agents Molecular Profile Target Setting Design Patients Primary Endpoint Results/States PYNNACLE [ 35 , 76 ] NCT04585750 Rezatapopt TP53 Y220C mutation TP53 - Phase I/II, Non-Randomized, Open label Solid tumor including ovarian cancer Response rate Interim investigator-assessed ORR=46% (95% CI, 31-61) DCO: 4 Sep 2025, recruitment and follow-up ongoing SORAYA [ 77 ] NCT04296890 Mirvetuximab soravtansine-gynx FRα-positive FRα Platinum-resistant recurrent Phase II, Single Group Assignment, Open label HGSOC ORR ORR=32.4% (95% CI, 23.6 to 42.2) DCO: 16 Nov 2021 MIRASOL [ 78 ] NCT04209855 Mirvetuximab soravtansine-gynx vs Chemotherapy FRα-positive FRα Platinum-resistant recurrent Phase III, Randomized, Open Label HGSOC PFS OS: 16.46 vs 12.75 mo; HR=0.67 (95% CI, 0.50-0.89) PFS: 5.62 vs 3.98 mo, HR=0.65 (95% CI, 0.52-0.81) DCO: 6 Mar 2023 16-724 [ 79 , 80 ] NCT02595892 Berzosertib + Gemcitabine vs Gemcitabine RS high / RS low ATM high / ATM low as post hoc analysis ATR Platinum-resistant recurrent Phase II, Randomized, Open Label HGSOC PFS PFS: 5.3 vs 3.4 mo, HR=0.57 (90% CI, 0.33-0.98) DCO: 21 Feb 2020 CAPRI [ 81 ] NCT03462342 Olaparib + Ceralasertib Cohort A/B: no limit of HR, Cohort C: HRD PARP1/2, ATR Platinum-sensitive / resistant recurrent Phase II, Non-Randomized, Open Label HGSOC Toxicity and efficacy Cohort B: No objective responses, but well-tolerated, DCO: 10 Jun 2020 NCI-2014-00620 [ 82 ] NCT02151292 Adavosertib + Gemcitabine vs Gemcitabine TP53 analysis, replication stress, DNA damage response pathway alteration as review WEE1 Platinum-resistant recurrent Phase II, Randomized, Single-blind All histologic subtypes of EOC PFS HGSOC PFS: 4.6 mo vs 3.0 mo HR=0.55 (95% CI, 0.35-0.90) HGSOC OS: 11.4 mo vs 7.2 mo HR=0.56 (95% CI, 0.35-0.91) DCO: Not reported 1775-004 [ 83 ] NCT01357161 Adavosertib vs placebo, plus carboplatin and paclitaxel TP53 WEE1 Platinum-sensitive disease Phase II, Randomized, Double-blind non-low grade, non-borderline ovarian, fallopian tube, or primary peritoneal cancer ePFS ePFS: 7.9 mo vs 7.3 mo, HR=0.63 (95% CI, 0.38-1.06) DCO: Not reported 140156 [ 84 ] NCT02203513 Prexasertib Ovarian cancer cohort: BRCA1/2wt CCNE1 status as post hoc analysis CHK1 Platinum-resistant recurrent Phase II, Single Group Assignment, Open Label HGSOC ORR ORR 30.8% (12/39; 95% CI, 17.0-47.6) DCO: Not reported 16712 [ 85 ] NCT03414047 Prexasertib BRCAmut, BRCAwt CHK1 Platinum resistant/refractory disease phase II, Non-Randomized, Open Label HGSOC ORR Platinum resistant cohort ORR=12.1% Independent of BRCA status DCO: Not reported PLX124-01 [ 86 ] NCT03297424 PLX2853 - BRD4 - Phase 1b/2a, Single Group Assignment, Open Label Solid tumor including 11 ovarian cancer cases safety, PK and RP2D RP2D= 80 mg/d DCO: 2 Feb 2021 REFMAL 495 [ 75 ] NCT03205176 AZD5153 + Olaparib vs AZD5153 BRCA/HRD status as post hoc analysis BRD4, PARP1/2 - Phase I, Sequential Assignment, Open Label Solid tumor including 14 ovarian cancer cases DLT RP2D=30 mg/d or 15 mg twice daily (mono) RP2D=10 mg/d (combination) DCO: Not reported PFS, Progression-Free Survival; ePFS, PFS by enhanced RECIST v1.1; ORR, Objective Response Rate; RP2D, Recommended Phase II Dose; PK, Pharmacokinetics; DLT, Dose-Limiting Toxicity; HR, Hazard Ratio; CI, confidence interval; DCO, data cut-off. Data cut-off dates are shown when reported in the original trial publications Representative Non-PARPi Targeted Therapy Clinical Trials in HGSOC Interim investigator-assessed ORR=46% (95% CI, 31-61) DCO: 4 Sep 2025, recruitment and follow-up ongoing ORR=32.4% (95% CI, 23.6 to 42.2) DCO: 16 Nov 2021 Mirvetuximab soravtansine-gynx vs Chemotherapy OS: 16.46 vs 12.75 mo; HR=0.67 (95% CI, 0.50-0.89) PFS: 5.62 vs 3.98 mo, HR=0.65 (95% CI, 0.52-0.81) DCO: 6 Mar 2023 RS high / RS low ATM high / ATM low as post hoc analysis PFS: 5.3 vs 3.4 mo, HR=0.57 (90% CI, 0.33-0.98) DCO: 21 Feb 2020 Olaparib + Ceralasertib Cohort A/B: no limit of HR, Cohort C: HRD PARP1/2, ATR Cohort B: No objective responses, but well-tolerated, DCO: 10 Jun 2020 Adavosertib + Gemcitabine vs Gemcitabine Phase II, Randomized, Single-blind HGSOC PFS: 4.6 mo vs 3.0 mo HR=0.55 (95% CI, 0.35-0.90) HGSOC OS: 11.4 mo vs 7.2 mo HR=0.56 (95% CI, 0.35-0.91) DCO: Not reported Adavosertib vs placebo, plus carboplatin and paclitaxel Phase II, Randomized, Double-blind ePFS: 7.9 mo vs 7.3 mo, HR=0.63 (95% CI, 0.38-1.06) DCO: Not reported Ovarian cancer cohort: BRCA1/2wt CCNE1 status as post hoc analysis Phase II, Single Group Assignment, Open Label ORR 30.8% (12/39; 95% CI, 17.0-47.6) DCO: Not reported phase II, Non-Randomized, Open Label Platinum resistant cohort ORR=12.1% Independent of BRCA status DCO: Not reported RP2D= 80 mg/d DCO: 2 Feb 2021 AZD5153 + Olaparib vs AZD5153 BRD4, PARP1/2 RP2D=30 mg/d or 15 mg twice daily (mono) RP2D=10 mg/d (combination) DCO: Not reported PFS, Progression-Free Survival; ePFS, PFS by enhanced RECIST v1.1; ORR, Objective Response Rate; RP2D, Recommended Phase II Dose; PK, Pharmacokinetics; DLT, Dose-Limiting Toxicity; HR, Hazard Ratio; CI, confidence interval; DCO, data cut-off. Data cut-off dates are shown when reported in the original trial publications DNA damage repair is intrinsically connected with cell-cycle checkpoint control, as many DDR proteins exert their function by activating specific checkpoints to induce cell-cycle arrest and provide sufficient time for lesion resolution. Cell-cycle checkpoints can be divided into two groups: DNA damage checkpoint and DNA replication stress checkpoint [ 45 ]. The DNA damage checkpoint operates through the ATM-CHK2-p53 pathway, which governs the G1/S transition in normal cells. This axis responds rapidly to irreparable DNA lesions and directs cells toward quiescence, senescence, or irreversible cell death. Mutations or functional attenuation of this pathway are common in tumor cells, contributing to checkpoint escape and oncogenic progression. The DNA replication stress checkpoint functions primarily through the ATR-CHK1-WEE1 axis, which regulates the G2/M checkpoint by stabilizing replication forks, alleviating the heightened replication stress characteristic of rapidly proliferating tumor cells, and preventing replication catastrophe and subsequent cell death [ 87 , 88 ]. From the perspective of cell-cycle governance, the ATR-CHK1-WEE1 pathway plays a tumor-suppressive role. Consequently, pharmacologic inhibition of this axis has emerged as a rational therapeutic strategy in tumors exhibiting high replication stress, among which includes HGSOC (Fig. 2 c). DNA damage activates ataxia telangiectasia mutated (ATM), which phosphorylates CHK2 and subsequently activates p53 by disrupting its interaction with the E3 ubiquitin ligase MDM2, thereby inhibiting MDM2-mediated p53 degradation [ 40 , 87 ] and inducing G1/S checkpoint arrest. In HGSOC, TP53 mutations occur in 96% of cases, with additional ATM mutations also documented [ 9 ]. The common TP53 and/or ATM mutations in HGSOC result in defective G1/S checkpoint control that drives elevated replication stress and premature S-phase entry, thereby compelling tumor cells to depend on the G2/M checkpoint to tolerate excessive replication stress and evade cell death. Preclinical studies have shown that ATR inhibitors selectively kill ATM- or p53-deficient cancer cells [ 89 ], and clinical trials exploring synthetic lethality in HGSOC via ATM-CHK2-p53 pathway disruption have consequently focused on targeting the G2/M checkpoint with ATR, CHK1, or WEE1 inhibitors, as summarized in Table 3 . A 2021 Phase II clinical trial ( NCT02595892 ) first demonstrated that the ATR inhibitor (ATRi) berzosertib, when combined with gemcitabine, improved patients’ PFS (22.9w vs 14.7w, HR=0.57) [ 79 ], and subsequent studies indicated that its efficacy was associated with replication stress [ 80 ]. In the CAPRI study, which comprised three cohorts, Cohort B reported treatment responses in platinum-resistant HGSOC (unselected for HRD status) receiving the combination of the ATRi ceralasertib and the PARPi olaparib. Despite limited sample size, a positive association between HRD status and clinical efficacy was observed, as BRCA1-mutant patients exhibited longer mPFS than the overall cohort (8.2 mo vs 4.2 mo), while subgroup analyses stratified by BRCA1/2 mutation status are still pending [ 81 ]. WEE1 inhibitor (WEE1i) has also been demonstrated significant clinical efficacy in multiple trials, as ovarian cancer patients harboring TP53 mutations showed superior outcomes with WEE1i compared with placebo, although both clinical benefits and adverse effects varied according to specific TP53 mutation subtypes [ 83 ]. In 2021, a Phase 2 clinical trial evaluating the first-line combination of adavosertib and gemcitabine demonstrated that WEE1i combination therapy provided greater benefit than gemcitabine monotherapy in platinum-resistant HGSOC patients (mPFS: 4.6 mo vs 3.0 mo, HR=0.55) (NCT02151292) [ 82 ]. Subsequent WES analysis revealed that 97% of enrolled patients carried TP53 mutations, and that the combination therapy exhibited enhanced efficacy in individuals with homologous recombination deficiency, thereby suggesting that these molecular subtypes could serve as predictive biomarkers for adavosertib-gemcitabine combination therapy. Additionally, the study incorporated a non-high-grade serous ovarian cancer exploratory single-arm cohort, in which response rates to the combination regimen were lower than those observed in the HGSOC cohort (partial response rate: 16% vs. 23%), although further validation of these findings is required [ 82 ]. A multi-cohort Phase 2 clinical trial evaluated the efficacy of CHK1/2 inhibitors across patients with BRCA1/2 mutation-associated breast or ovarian cancer, triple-negative breast cancer, BRCAwt HGSOC, and metastatic castration-resistant prostate cancer, demonstrating an objective response rate of 30.8% for CHK1/2 inhibitor monotherapy (prexasertib) in BRCAwt HGSOC [ 84 ]. However, CHK1 inhibition efficacy exhibited no significant correlation with clinical features, BRCA status, or prior treatments, highlighting the urgent need for clinically meaningful biomarkers predictive of CHK1 responsiveness [ 85 ]. The above ATR/CHK1/WEE1 pathway-targeted strategies collectively underscore that integrating histotype and molecular subtypes can enhance therapeutic outcomes. CCNE1 encodes Cyclin E1, which partners with CDK2 to drive the G1/S transition, and CCNE1 amplification is observed in approximately 20% of HGSOC cases. In this review, CCNE1 amplification refers to high-level copy-number amplification, defined as copy number greater than 5 [ 90 ]. CCNE1 amplification accelerates premature S-phase entry, exacerbates DNA damage, and further destabilizes G1/S checkpoint control which has already been compromised by TP53 mutations, thereby increasing dependency on the G2/M checkpoint. Consequently, CCNE1 has emerged as a biomarker predictive of G2/M checkpoint inhibitor (ATRi/CHK1i/WEE1i) efficacy. Preclinical studies have demonstrated that low-dose WEE1i-ATRi combination therapy synergistically could reduce HGSOC cell viability and clonogenic potential in a CCNE1-dependent manner [ 90 ]. The NCT03253679 trial recruited patients with advanced or refractory solid tumors harboring CCNE1 amplification to evaluate adavosertib (WEE1i) efficacy, and notably, among the 14 ovarian cancer patients in this basket trial, encouraging responses were observed (ORR=36%) [ 91 ], suggesting considerable clinical potential for CCNE1 amplification as a predictive marker. Beyond intracellular vulnerabilities related to DNA damage repair and cell-cycle checkpoint dependence, cell-surface antigen expression provides another clinically validated route for biomarker-directed therapy in HGSOC and related epithelial ovarian cancers. FRα, encoded by FOLR1, is a cell-surface folate-binding protein involved in folate transport. Its restricted expression in most normal adult tissues, frequent overexpression in epithelial malignancies, and capacity for receptor-mediated internalization provide the biological rationale for FRα-directed drug delivery [ 92 ]. In ovarian cancer, immunohistochemical analyses have shown frequent FRα expression in epithelial ovarian tumors, supporting its development as a therapeutic target [ 93 ]. Mirvetuximab soravtansine-gynx, also known as IMGN853, was developed on this basis as an FRα-directed antibody-drug conjugate that links an anti-FRα antibody to the maytansinoid microtubule inhibitor DM4. Preclinical studies demonstrated that IMGN853 selectively binds FRα-positive tumor cells, undergoes internalization, releases DM4 intracellularly, disrupts microtubule dynamics, and induces cell-cycle arrest and tumor-cell death; antitumor activity was observed in FRα expressing xenograft models, including ovarian cancer models [ 94 , 95 ]. This mechanism differs from PARP inhibitors or ATR/CHK1/WEE1 inhibitors, as mirvetuximab soravtansine relies on tumor antigen expression rather than DNA repair deficiency for biomarker selection. The clinical relevance of FRα selection was first supported by the phase II SORAYA trial, in which mirvetuximab soravtansine showed an objective response rate of 32.4% and a median duration of response of 6.9 months in patients with FRα-high, platinum-resistant ovarian cancer previously treated with bevacizumab [ 77 ]. This benefit was subsequently confirmed in the phase III MIRASOL trial ( NCT04209855 ), which enrolled patients with FRα-positive, platinum-resistant high-grade serous ovarian cancer after one to three prior systemic treatment regimens and randomized them to mirvetuximab soravtansine or investigator’s choice chemotherapy. Mirvetuximab soravtansine significantly improved overall survival (16.46 mo vs 12.75 mo; HR=0.67), progression-free survival (5.62 mo vs 3.98 mo; HR=0.65), and objective response rate (42.3% vs 15.9%) [ 78 ]. Although the pivotal MIRASOL evidence was generated in high-grade serous ovarian cancer, the FDA full approval granted on March 22, 2024 applies to adults with FRα-positive, platinum-resistant epithelial ovarian, fallopian tube, or primary peritoneal cancer after one to three prior systemic treatment regimens, without restricting histotypes. This approval establishes FRα expression as one of the most clinically mature non-PARPi biomarkers in epithelial ovarian cancer.

Ovarian cancer, characterized by its insidious onset, histological diversity, and frequent late-stage diagnosis, remains a major global health burden among women. According to GLOBOCAN 2024 estimates, ovarian cancer ranked eighth in both cancer incidence and cancer mortality among women worldwide, with an estimated 330,731 new cases and 203,850 deaths globally. The corresponding age-standardized incidence and mortality rates were 6.5 and 3.8 per 100,000 women, respectively [ 1 , 2 ]. Generally, primary malignant ovarian tumors consist of three main types: epithelial ovarian cancer (EOC, accounting for approximately 90%), sex cord stromal tumors, and germ cell tumors. EOC comprises several distinct subtypes, including high-grade serous ovarian carcinoma (HGSOC, 63.4%), low-grade serous ovarian carcinoma (LGSOC, 2.5%), clear cell carcinoma (9.6%), endometrioid carcinoma (9.9%), and mucinous carcinoma (9.4%), with rarer histotypes accounting for the remainder [ 3 , 4 ]. At the microscopic level, different histotypes of EOC exhibit distinct molecular genetic profiles [ 5 ] (Fig. 1 ). High-grade serous ovarian carcinoma (HGSOC) is characterized by extensive genomic instability, with the most frequent molecular alterations including TP53 mutations (present in nearly all cases, 96%), BRCA1/2 germline or somatic mutations (~21%), CCNE1 amplification (~20%), and BRD4 amplification (18%) [ 6 – 10 ]. Moreover, folate receptor alpha (FRα)/FOLR1 expression is reported in approximately 44% of HGSOC cases [ 11 ]. Therapeutic strategies targeting these alterations have entered clinical practice or are under clinical investigation. In addition, some HGSOC cases also show NF1 and RB1 deletions [ 9 ]. Low-grade serous ovarian carcinoma (LGSOC) displays relative genomic stability compared with HGSOC and is distinguished by mutations in MAPK pathway genes [ 12 ], including KRAS (33.0%), NRAS (11.2%), and BRAF (8.4%) [ 13 , 14 ]. Besides, CDK4/6 has been explored as a potential therapeutic target in LGSOC, particularly in combination with endocrine therapy [ 15 ]. Ovarian clear cell carcinoma (OCCC) and endometrioid ovarian carcinoma (EnOC), collectively referred to as endometriosis-associated ovarian cancer (EAOC), share similar molecular profiles, such as ARID1A mutations (46%-57%) [ 16 , 17 ] and PIK3CA activation (40%) [ 18 ]. Research into novel therapeutic targets for EAOC remains largely preclinical, with only a limited number of clinical trials currently available. Clinical studies exploring new therapeutic targets for mucinous ovarian carcinoma (MOC) remain limited, and some previous attempts were terminated or yielded disappointing outcomes [ 19 , 20 ]. Fig. 1 Molecular alterations and downstream biological consequences across major epithelial ovarian cancer histotypes. a) In HGSOC, defects in DNA damage repair and cell-cycle regulation drive genomic instability and unrepaired DNA damage, while FRα expression provides a cell-surface biomarker for FRα-directed antibody-drug conjugate delivery. b) LGSOC is characterized by MAPK-pathway activation, together with ER/PR signaling, sustaining hormone-responsive growth. c) EAOC commonly exhibits ARID1A loss, leading to chromatin remodeling defects, and may also show mismatch repair deficiency (MMRd), which is associated with increased mutational burden. d) MOC shows ERBB2 amplification, which may increase the potential applicability of antibody-drug-conjugate-based therapeutic approaches. These histotype-specific perturbations underscore the diverse targeting strategies underlying epithelial ovarian cancer. Created with BioGDP.com [ 21 ] Molecular alterations and downstream biological consequences across major epithelial ovarian cancer histotypes. a) In HGSOC, defects in DNA damage repair and cell-cycle regulation drive genomic instability and unrepaired DNA damage, while FRα expression provides a cell-surface biomarker for FRα-directed antibody-drug conjugate delivery. b) LGSOC is characterized by MAPK-pathway activation, together with ER/PR signaling, sustaining hormone-responsive growth. c) EAOC commonly exhibits ARID1A loss, leading to chromatin remodeling defects, and may also show mismatch repair deficiency (MMRd), which is associated with increased mutational burden. d) MOC shows ERBB2 amplification, which may increase the potential applicability of antibody-drug-conjugate-based therapeutic approaches. These histotype-specific perturbations underscore the diverse targeting strategies underlying epithelial ovarian cancer. Created with BioGDP.com [ 21 ] With increasing recognition of ovarian cancer as a highly heterogeneous disease, molecularly targeted therapies such as PARP inhibitors have been incorporated into first-line maintenance treatment. However, current clinical practice generally lacks stratification by histotypes or molecular classifications to guide targeted therapeutic strategies [ 22 ]. The successful application of molecular classification in breast, colorectal, and endometrial cancers has demonstrated the potential efficacy of molecular subtype directed therapies [ 23 – 27 ]. In this review, molecular characteristics are integrated with histotype classification to summarize potential molecular targets and corresponding targeted strategies beyond PARP inhibitors, across different ovarian cancer histotypes, while also updating relevant clinical trial evidence. The goal is to explore novel targeted therapeutic approaches for ovarian cancer from a histotype-oriented perspective.