Morphological and molecular characterization of epithelial ovarian cancer: SEOM-SEAP consensus guidelines for biomarker integration in clinical practice

Abstract

Epithelial ovarian cancer (OC) is the second leading cause of death among gynecologic malignancies and exhibits marked histological and molecular heterogeneity. Advances in tumor biology and targeted therapies have increased the need for accurate biomarker characterization. This updated consensus by the Spanish Society of Medical Oncology (SEOM) and the Spanish Society of Pathology (SEAP) reviews key morphological and molecular features for diagnosis and providing recommendations for biomarker assessment across major OC subtypes. The document addresses classical serum biomarkers, prognostic tools, and genomic alterations with therapeutic implications, including homologous recombination deficiency (HRD), BRCA1 and BRCA2 mutations, mismatch repair status, as well as other established and emerging biomarkers. Emphasis is placed on optimal sample selection, validated techniques, and multidisciplinary integration is required to ensure diagnostic quality. These recommendations aim to optimize prognostic stratification and guide personalized treatment strategies, ultimately improving clinical outcomes in ovarian cancer.

Introduction

Epithelial ovarian cancer (OC) is the second leading cause of cancer-related death among gynecological malignancies. It is also the eighth most common cancer diagnosis and cause of cancer death in women [1].

It is estimated that 1 in 78 women (1.3%) will develop OC at some point during her lifetime [2]. Although high-grade serous ovarian carcinoma (HGSOC) is the most common histological type, ovarian tumors comprise a group of heterogeneous neoplasms and diagnosis can be difficult for pathologists. Immunohistochemistry and other molecular tests are used to classify ovarian tumors, although histology remains the mainstay in diagnosis [3].

Five main types of OCs are currently identified: high-grade serous carcinomas (70%), endometrioid carcinomas (10%), clear cell carcinomas (10%), mucinous carcinomas (3%), and low-grade serous carcinomas (< 5%) [3]. This consensus focuses on these five main types and briefly mentions other rarer ovarian carcinomas.

Advances in molecular biology and diagnostic techniques in recent years have enabled a better understanding of the key molecular alterations involved in the development of these tumors. At the same time, the treatment of OC has evolved, underscoring the need for high-quality pathology diagnosis in order to ensure the best treatment options for each patient.

The purpose of this revised consensus guideline of the Spanish Society of Medical Oncology (SEOM) and the Spanish Society of Pathological Anatomy (SEAP) is to review the key morphological and molecular features for diagnosis and the main recommendations for determining molecular alterations in each type of ovarian cancer. This consensus guideline is an update to the one published in 2017 [4].

The role of current and emerging promising biomarkers is also discussed.

Histopathological classification of epithelial ovarian cancer

High-grade serous carcinoma

High-grade serous ovarian carcinoma (HGSOC) is the most common (70%) and lethal malignancy among ovarian tumors [5]. Approximately 80% of patients are diagnosed at an advanced stage. HGSOC typically arises from the tubal-type epithelium, most commonly in the fimbriae of the tube, though it can also arise from the ovarian surface epithelium [3].

These tumors are typically bilateral (66%) and tend to be large, albeit smaller than other forms of OC. Macroscopically, HGSOC is characterized by a solid-papillary mass with frequent areas of extensive necrosis and fluid-filled cysts. Microscopically, HGSOCs are heterogeneous tumors presenting solid, papillary, glandular or cribriform patterns, with common necrosis. The tumor cells are of intermediate size, although scattered atypical giant cells with large nucleoli may also be observed. Mitotic activity is high, often exceeding 12 mitoses per 10 high-power fields (HPF). HGSOCs present positive immunoreactivity for several markers, including Cytokeratin 7 (CK7), Paired box 8 (PAX8) (95%), Wilms tumor 1 (WT1) (90%), Estrogen receptor (ER) (80%), Progesterone receptor (PR) (30%), p16^INK4a (≥ 50%), and an abnormal pattern of p53 expression (mutated pattern).

Nearly all HGSOCs harbor deleterious TP53 mutations, driving chromosomal instability and complex copy-number alterations [6]. HGSC is also associated with germline or somatic mutations in BRCA1/2 genes, which encode crucial proteins for DNA repair in the homologous recombination repair (HRR) pathway. Mutations in these genes, observed at high frequency (about 50%) in HGSOC patients, can lead to a deficient HRR pathway, recognized as homologous recombination deficiency (HRD) [7].

HGSOC includes a morphologic variant characterized at microscopic appearance by solid, pseudo-endometrioid and transitional-like architectural-features, collectively identified as the SET pattern [8]. SET tumors exhibit a higher number of tumor-infiltrating lymphocytes, elevated intratumoral heterogeneity, and a greater mitotic index. Approximately 90% of HGSOCs with the SET pattern are considered HRD tumors. Identifying the SET variant of HGSC may help select patients most likely to benefit from PARP inhibitor therapy [8].

Endometrioid carcinoma

Endometrioid ovarian carcinoma (EOC) accounts for 10% of all OCs. It is more commonly diagnosed in perimenopausal women and frequently identified at an early stage. Endometriosis is associated with an increased risk for developing EOC with ipsilateral ovarian or pelvic endometriosis being observed in more than 40% of cases [9]. EOCs are usually unilateral in over 80% of cases and can be associated with synchronous carcinoma of the endometrium in 10–20% of the cases [3, 10]. Microscopically, EOCs exhibit near-identical characteristics to endometrioid carcinoma of the uterine corpus. Most EOCs are low-grade adenocarcinomas; however, in about 10% of cases, a solid growth pattern predominates, often concurrent with geographic necrosis [11]. Squamous differentiation occurs in about 50% of cases, while rarer forms of differentiation include mucinous, secretory, fusiform, ciliated, oxyphilic, sex-cord-like, and clear cell changes. Immunohistochemically, EOCs are positive for vimentin, CK7 (97%), PAX8 (82%), ER (77%), and PR (67%) and typically negative for WT1 (10%), Napsin A (3–8%), ɑ-inhibin and calretinin [12].

Seromucinous carcinomas are considered a subtype of EOC, as molecular studies suggest that they are morphologic variants of EOC [3, 13]. The most prevalent molecular alterations in EOC involve key signaling pathways, including mutations in the WNT/β-catenin pathway (CTNNB1 mutations observed in 43–53% of cases) which is frequently affected. Other alterations include mutations in the PI3K pathway (PI3KCA mutations in 40–43% and PTEN mutations in 17–29%), the mitogen-activated protein kinase (MAPK) pathway (KRAS mutations in 26–33%), and the SWI/SNF complex (ARID1A mutations in 30–36%). CTNNB1 mutations are associated with squamous differentiation, low tumor grade, and favorable outcomes [3, 14]. Similar to endometrial endometrioid carcinoma, EOC is classified into four molecular subtypes according to The Cancer Genome Atlas (TCGA): ultramutated (due to POLE exonuclease domain mutations; 5%), hypermutated (due to mismatch repair protein deficiencies, MMRd; 13%), TP53 mutated (9–13%), and subtypes with no specific molecular profile (NSMP; 69–73%) [11].

Clear cell carcinoma

Clear cell ovarian carcinomas (CCOCs) constitute approximately 10% of OCs which are frequently diagnosed at early stages, often presenting as a unilateral ovarian mass. These tumors are strongly correlated with endometriosis, present in 50–74% of cases [3, 15]. Patients with this condition tend to have a more favorable prognosis [3, 15]. CCOCs exhibit a combination of papillary, tubulocystic, and solid growth patterns, and their cysts and glands are lined by cuboidal, hobnail, or flattened cells with clear or eosinophilic cytoplasm [15]. Distinctive morphological features include multiple complex papillae, dense hyaline basement membrane material, and hyaline bodies [3]. Stromal hyalinization and myxoid stroma are frequent. Immunohistochemically, CCOCs are typically positive for PAX8, Napsin A, and Hepatocyte Nuclear Factor 1-beta (HNF1β), and negative for WT1, ER and PR.

The most common and significant molecular alterations in CCOCs include inactivating mutations of ARID1A (50%), a tumor suppressor and key component of the SWI/SNF chromatin remodeling complex, and activating mutations of PIK3CA (50%), which impact the PI3K/AKT signaling pathway. These mutations co-occur in 20–56% of CCOCs and are believed to arise early in tumorigenesis, as they are also detected in endometriosis [15]. Recent evidence suggests that ARID1A mutations may improve responses to cancer immunotherapy due to enhanced tumor immunogenicity [16]. Tumors harboring PIK3CA mutations tend to exhibit histological features, such as hyalinized or mucoid stroma, and are associated with an elevated risk of paraneoplastic thromboembolism [17]. Additionally, hepatocyte nuclear factor-1β (HNF-1β) is consistently upregulated in CCOCs. Reported genetic alterations in CCOCs include TERT promoter mutations (16%), KRAS mutations (10%), and TP53 mutations (< 10%). Microsatellite instability (MSI) is observed in a small percentage of CCOCs (< 10%), a subgroup that may exhibit favorable clinical outcomes despite advanced disease, potentially owing to tumor immunogenicity and responsiveness to immunotherapy. Ovarian CCOCs exhibit a unique epigenetic profile that includes hypomethylation of promoter regions in multiple genes within the HNF1 pathway and hypermethylation of promoter regions associated with the ER alpha (ERα) pathway [3].

Low-grade serous carcinoma

Low-grade serous ovarian carcinomas (LGSOCs) account for less than 5% of OCs. These tumors typically arise from the transformation of a serous borderline tumor (SBT) into an LGSOC. However, this progression is rare, affecting less than 10% of patients, and usually develops over a long period of time. A non-invasive SBT component is often observed alongside LGSOC. The diagnosis of “microinvasive” LGSOC, defined as ovarian stromal invasion of less than 5mm, does not correlate with a poor prognosis and is managed similarly to an SBT [18]. According to the 2020 WHO nomenclature, the presence of invasive peritoneal implants in an SBT warrants a diagnosis of LGSOC [11]. However, there is no consensus on whether these implants should be considered equivalent to metastatic LGSOCs [18]. LGSOCs are characterized by tumor cells with uniform, small nuclei arranged in small nests, glands, papillae, or micropapillae. These are often embedded in a variably hyalinized stroma that may contain psammoma bodies. A key distinguishing feature of LGSOCs compared to HGSOCs is their uniform nuclear morphology and low Ki-67 labeling index. Immunohistochemical profiling reveals diffuse positivity for CK7, WT1, PAX8, and ER. Meanwhile, p16 (CDKN2A gene) shows patchy expression and p53 typically exhibits a wild-type staining pattern. LGSOCs are characterized by alterations in the MAPK pathway, with mutations in the KRAS and BRAF genes being the most prevalent. Additional mutations including those in the NRAS, ERBB2, PIK3CA, CDKN2A/2B, USP9X, and EIF1AX genes have also been documented in these tumors. Remarkably, KRAS and BRAF mutations are mutually exclusive and are believed to occur early in the development of SBTs and LGSOCs. Although BRAF mutations are less frequent in LGSOCs than in SBTs, their presence has been associated with early-stage disease and a more favorable prognosis [19]. Conversely, KRAS mutations in SBTs are associated with tumor recurrence, suggesting their potential as biomarkers for recurrence risk [20].

Mucinous carcinoma

Mucinous ovarian carcinomas (MOCs) represent only 3–4% of primary epithelial ovarian carcinomas, even though mucinous tumors comprise 10–15% of all ovarian tumors, the majority being benign or borderline [3]. MOC cells may resemble those of the gastric pylorus, intestine, or endocervix. However, most MOCs exhibit an epithelium that is more akin to the gastric foveolar or pancreaticobiliary types rather than the intestinal-type [3, 21]. A spectrum of architectural and cytological changes, including benign, borderline, non-invasive carcinoma, and invasive components, may coexist within a single MOC tumor [3, 16, 21]. This supports the idea that MOCs evolve progressively from borderline tumor precursors to high-grade invasive MOCs [22]. Microscopically, MOCs can exhibit two distinct growth patterns: the expansile type, which is associated with a more favorable prognosis and is characterized by complex malignant glands without obvious stromal invasion, and the infiltrative type, which shows evident stromal invasion and is frequently associated with a desmoplastic stromal reaction [3, 16, 21, 23]. MOCs typically exhibit a diffuse and strong immunoreactivity for CK7, while showing weak and focal reactions for CK20, CEA, and CDX2. MOCs are generally negative for WT1 and PAX8, or show weak and focal expression of these markers. MOCs are also negative for ER, PR, and vimentin, while p53 may display either wild-type or mutation-type staining, and p16 typically shows a non-block type staining pattern.

The main differential diagnosis for primary MOCs are metastatic mucinous carcinomas (MMCs). Primary MOCs are typically unilateral, large, and confined to the ovary. In contrast, MMCs are usually smaller and bilateral. They often involve the ovarian surface and can result in pseudomyxoma peritonei [3]. MMCs commonly exhibit widespread infiltrative invasion. Genetic alterations in MOCs often include CDKN2A copy number loss or mutations, which are observed in 76% of cases. KRAS and TP53 mutations follow, each occurring in 64% of cases. ERBB2 amplification is found in 26% of cases, while mutations in RNF43, BRAF, PIK3CA, and ARID1A are identified in 8–12% of cases [21, 22]. Benign tumors are typically initiated by genetic events involving either KRAS or CDKN2A [21, 22] (Table 1).

Table 1.

Clinicopathological and molecular characteristics of ovarian cancer subtypes

	High-grade serous	Endometrioid	Clear cell	Low-grade serous	Mucinous
Precursor tissue/lesion	Fallopian tube/tubal neometaplasia	Endometriosis; adenofibroma	Endometriosis; adenofibroma	Serous borderline tumor	Mucinous borderline tumor
IHC Positive	CK7, PAX8 (95%), WT1 (90%), ER (80%), TP53 (mutated pattern), P16 (> = 50%)	CK7 (97%), PAX8 (92%), ER (77%), PR (67%)	PAX8, Nap A, HNF1B	CK7, WT1, PAX8, ER	CK7, CK20, CEA, CDX2, TP53 (mutated pattern)
IHC Negative	PR (+ 30%)	WT1 (+ 10%), Nap A (+ 3–8%), Alfa-inhibina, calretina, TP53 (wildtype 85%)	WT1, ER, PR, TP53 (wildtype 88%)	P16 (patchy), TP53 (wildtype)	WT1, PAX8, ER, PR, TP53 (wild-type)
Main molecular abnormalities	TP53, BRCA-HRD, HER2	Wnt/B-catenin, MSI, PIK3KA, ARID1A, PTEN	ARID1A, PTEN, PIK3KA	KRAS, BRAF	CDKN2A, KRAS, ERBB2, TP53
Stage at diagnosis	Advanced	Early	Early	Early/advanced	Early
Response to chemotherapy	High		Low	Intermediate	Low
Prognosis	Poor (all stages)	Favorable (early stage) Intermediate (advanced stage)	Intermediate (all stages)	Favorable (early stage) Poor (advanced stage)	Favorable (early stage) Poor (advanced stage)

Other carcinomas

The latest WHO classification of female genital tumors [11] recognizes additional rare carcinomas whose identification is important due to their aggressive behavior or diagnostic overlap with common ovarian malignancies.

Mesonephric-like adenocarcinoma of the ovary typically presents as a unilateral, stage I tumor that shares the same morphology as other mesonephric carcinomas exhibiting immunoreactivity for GATA3, TTF-1, PAX8, and CD10. It is negative for ER, PR, and WT1, and displays a wild-type p53 pattern.

Undifferentiated carcinoma is a malignant epithelial neoplasm without evidence of specific differentiation. In contrast, dedifferentiated carcinoma consists of a differentiated carcinoma and an undifferentiated component. Both carcinomas share morphological similarities with their endometrial counterparts, exhibit aggressive behavior, and typically present at advanced stages. They are immune-negative for ER, PR, and E-cadherin. MMRd and mutations in the chromatin remodeling genes ARID1A/B and SMARCA4/A2/B have been identified.

Ovarian carcinosarcoma is a biphasic tumor consisting of a high-grade carcinoma, typically HGSOC, and a sarcomatous component. TP53 mutations are common, as are other genetic alterations shared with its uterine counterpart. Mixed carcinoma of the ovary is defined as an ovarian carcinoma composed of two or more distinct histologic components. The most common subtype is a combination of endometrioid carcinoma (EOC) and clear cell carcinoma (CCOC) [24, 25].

Serum biomarkers

Cancer antigen 125

Cancer antigen 125 (CA-125), a glycoprotein encoded by the MUC16 gene on chromosome 19, is secreted from the coelomic and müllerian epithelia into the bloodstream. The established upper limit for CA-125 is 35 U/mL in both premenopausal and postmenopausal patients [26]. Elevated CA-125 levels are suggestive of potential malignancies, with a 47% increase noted in early-stage OC and a more pronounced 80–90% rise in advanced stages [27]. However, it can be within the normal range in up to 20% of ovarian cancers [28]. The expression of CA-125 varies by histological subtype, typically higher in serous and endometrioid tumors, and lower in non-epithelial, clear cell, undifferentiated and mucinous OCs [12]. The specificity of CA-125 is relatively modest (generally 73–77%) due to elevations observed during pregnancy, the menstrual cycle, other malignancies, and other benign conditions like acute pelvic inflammation, uterine myoma, endometriosis, as well as other non-gynecological diseases such as cirrhosis, hepatitis, pancreatitis and congenital heart disease [29].

The CA-125 response criterion for evaluating chemotherapy response was initially defined by Rustin and later adapted by the Gynecological Cancer Intergroup (GCIG). It is defined as a 50% reduction in CA-125 levels maintained for at least 28 days [30].

Early detection of relapse

Elevations in CA-125 can detect recurrent cancer 2–5 months before clinical diagnosis [31, 32], with a specificity and sensitivity in this setting being 91.3% and 85.9%, respectively. According to the GCIG, relapsed OC is defined in terms of CA-125 by one of the following criteria [32]: (1) in patients whose CA-125 level was elevated before treatment and returned to normal, CA-125 must be elevated above twice the upper limit of normal; (2) in patients whose CA-125 level was elevated before treatment but did not return to normal, CA-125 must rise to twice the nadir or above; or (3) in patients whose CA-125 is within the normal range, it must show a rise to twice the upper limit of normal or above. In all cases, elevated CA-125 must be confirmed on at least two occasions at least one week apart, to reduce the false positive rate to 2%. However, one study showed that early initiation of treatment based on CA-125 elevation alone did not improve survival and instead led to a decline in quality of life compared to starting treatment after clinical or radiologic confirmation of recurrence [33]. This study was conducted before the results of studies confirmed a survival benefit with second-line cytoreductive surgery, so ESMO and ESGO continue to recommend CA-125 monitoring to detect recurrence with low tumor burden [34].

Roles in screening and diagnosis

The effect of OC screening with CA-125 on mortality has been assessed in two large randomized studies. The PLCO trial [35] enrolled over 78,000 women (aged 55–74 years) and randomized them to either a screening group with annual transvaginal ultrasound and CA-125 testing, or a usual care group. After a median follow-up of 15 years, no reduction in mortality was observed in the screening arm. Similarly, the UKCTOCS trial [36] randomized over 202,000 postmenopausal women to one of three groups: annual multimodal screening (MMS) using the ROCA algorithm with serial CA-125 measurements and further evaluation by transvaginal ultrasound (TVS) if elevated levels were detected; TVS alone; or no screening. No significant survival benefit was demonstrated.

In BRCA1/2 mutation carriers, screening with TVS, pelvic examination, and CA-125 has failed to detect early-stage OC [37]. Based on these findings, routine OC screening is not recommended for the general population. In patients at genetic risk, TVS and CA-125 may be considered, although the benefit remains controversial.

Human epididymis secretory protein

The human epididymis secretory protein (HE4) is a peptide protease inhibitor involved in the innate immune response of epithelial tissues. It is overexpressed in ovarian, endometrial, and certain lung cancers, including adenocarcinomas and squamous cell carcinomas, as well as breast cancer and mesothelioma. However, HE4 levels are typically not elevated in benign conditions such as endometriosis or pelvic inflammatory disease. In OC, HE4 levels often exceed 70 pmol/L, distinguishing cancer patients from healthy individuals or those with benign conditions. When combined with CA-125, HE4 significantly improves diagnostic sensitivity and specificity for differentiating malignant from benign disease. Notably, HE4 is more specific than CA-125 in premenopausal patients, although CA-125 shows higher sensitivity in this subgroup [38].

Several studies suggest that HE4 may serve as a predictive factor for platinum response, progression-free survival (PFS), overall survival (OS) and surgical outcome, and may be more effective than CA-125 in predicting OC recurrence. Further validation is needed to confirm HE4’s role in long-term OC management [39].

Multivariate index assays for OC evaluation

Several indices have been developed to improve OC diagnosis by combining multiple biomarkers. The Risk of Malignancy Index (RMI) is calculated by multiplying ultrasound results (U), CA-125 levels, and menopausal status (M). With a cut-off value of 200, RMI enhances both sensitivity and specificity compared to CA-125 alone [39].

The OVA1 assay is an FDA-approved multivariate index that incorporates CA-125, transthyretin, transferrin, beta-2 microglobulin, and apolipoprotein A-1 to assess the ovarian malignancy risk. OVA1 demonstrated higher sensitivity (92% vs 79%) and a better negative predictive value (97% vs 93%) than CA-125, enabling earlier detection of OC and improving the malignancy detection rate [40]. The Risk of Ovarian Malignancy Algorithm (ROMA) is a multivariate index that integrates serum CA-125, HE4, and menopausal status using a logistic regression model. Approved by the FDA in 2010, ROMA outperforms CA-125 or HE4 alone in predicting OC. A meta-analysis of 5954 cases revealed a sensitivity of 90%, specificity of 91%, positive predictive value (PPV) of 90%, and negative predictive value (NPV) of 93%, confirming the utility of ROMA in the assessment of OC risk [41].

Biomarkers to evaluate response to platinum-based chemotherapy

Chemotherapy response score

Böhm et al. [42] developed and validated the chemotherapy response score (CRS), a histological scoring system for assessing response to neoadjuvant chemotherapy in advanced tubo-ovarian high-grade serous carcinoma. The CRS uses a three-category system: 1) CRS1—minimal tumor response; 2) CRS2—moderate response, with identifiable residual neoplastic foci, and 3) CRS3—complete or near-complete response, with no or minimal residual scattered neoplastic cells (< 2 mm). The CRS is a simple, reproducible score system for pathologists with significant correlations to PFS and OS when applied to omental specimens. A significant prognostic difference was observed between CRS1-2 and CRS3. Based on these findings, the International Collaboration on Cancer Reporting (ICCR) and the College of American Pathologists (CAP) have recommended CRS for standardized reporting of histological responses to neoadjuvant chemotherapy in high-grade serous carcinoma [43]. More recently, CRS has been confirmed and further validated both in its three- and two-tiered forms, as a useful prognostic tool for guiding clinical decisions [44].

In relation to the pathological assessment of the omentum, when macroscopic tumor is present, representative sections are considered sufficient. In the absence of visible disease, 4 to 6 sections targeting the most abnormal-appearing areas are recommended to ensure an adequate evaluation of response to neoadjuvant chemotherapy [45].

CA-125 elimination rate constant K

The elimination rate constant K (KELIM) is a mathematically modeled kinetic parameter based on CA-125 clearance during systemic treatment. KELIM has demonstrated higher prognostic value for PFS and OS compared to the GCIG response criterion [46]. KELIM has been studied as a predictor of chemotherapy sensitivity following primary cytoreductive surgery in large phase III trials (AGO-OVAR 7, AGO-OVAR 9, and ICON-7) [47], as well as in the neoadjuvant setting as a prognostic factor for complete cytoreduction in the CHIVA trial [48]. Recently, KELIM has been further evaluated as a prognostic factor in the setting of first-line maintenance treatment. In the ICON 7 trial [49], high-risk patients with unfavorable KELIM had the highest survival benefit from bevacizumab, while the VELIA trial [50] showed that patients with favorable KELIM experienced a significant PFS benefit with veliparib. An online calculator (http://www.biomarker-kinetics.org/CA-125) [47] enables clinicians to enter chemotherapy cycle dates and CA-125 values from the first 100 days of chemotherapy. This tool classifies patients as having a favorable (≥ 1), intermediate (0.5–1), or unfavorable (< 0.5) KELIM. This kinetic parameter is being prospectively assessed in the ongoing NIRVANA-1 and AGO-OVAR 28/ENGOT-ov57 trials, which compare niraparib with or without bevacizumab in patients after complete primary surgery [51].

Homologous recombination deficiency

The HRR pathway is critical for maintaining genomic stability by repairing DNA double-strand breaks and resolving interstrand cross-links (ICL) [52]. These types of DNA damage can lead to mutations, chromosomal rearrangements and ultimately genomic instability which contributes to tumorigenesis. The HRR system includes a complex network of genes that orchestrates DNA repair processes. Key components of the HRR system include the tumor suppressors BRCA1 and BRCA2, which stand out as primary mediators of DNA repair and genomic stability. Additionally, genes such as ATM, BARD1, BRIP1, H2AX, MRE11, PALB2, RAD51, RAD51C/D, RPA, and Fanconi Anemia genes also contribute to the HRR pathway, supporting various aspects of the repair process [52]. Genomic instability caused by mutations in HRR-related genes leads to HRD, a recognized molecular biomarker for guiding the use of PARP inhibitor (PARPi) or platinum-based (Pt) chemotherapy. Tumors that do not exhibit HRD are considered homologous recombination proficient (HRP) [53–55].

BRCA1 and BRCA2 alterations in OC

HRD is associated with several tumor types in OC and is frequently observed in high-grade OC. About 13–15% of women diagnosed with OC harbor germline mutations in BRCA1 or BRCA2 with a lifetime risk of developing OC of 40% for BRCA1 carriers and 18% for BRCA2 carriers [56]. Mutation prevalence varies by histological subtype: 14.5% in high-grade serous ovarian cancer (HGSOC), 7.7% in endometrioid, 4.9% in CCC, and 12.3% in other histological types [57] [58].

In addition to germline mutations in BRCA1/2, somatic mutations in these genes can also result in loss of function. These somatic alterations have been reported in approximately 5–7% of OCs, and should therefore be considered in the comprehensive assessment of HRD in these tumors. Importantly, relying solely on somatic testing may miss up to 5% of germline mutations, highlighting the importance of germline testing in these patients [58].

Epigenetic alterations can also silence BRCA1/2. Promoter hypermethylation of BRCA1 is a common mechanism of loss of function, as is overexpression of the EMSY gene. Methylation can also silence other genes in the HRR pathway, such as RAD51C. These changes contribute to HRD and provide potential therapeutic targets for intervention [58, 59].

Methods to assess HRD status

In addition to mutations and epigenetic alterations in BRCA1/2, other biomarkers and methods have been introduced to assess HRD status in OC. Gene expression profiling, proteomics, and RNA analyses have been used to generate “BRCAness” signatures or genomic instability scores reflecting defects in the HRR pathway. These signatures may help to identify HRD in tumors. Furthermore, signatures of chromosomal instability based on single-nucleotide polymorphism (SNP) arrays have been evaluated as a method to measure “genomic scarring”. Functional assays have also been explored, including the evaluation of RAD51 accumulation at DNA damage sites, which serves as a recognized marker of HRR proficiency [58].

HRD can be detected in tumor samples by evaluating either the causes (e.g., genetic and epigenetic events) or the consequences (e.g., molecular phenotype) of genomic instability, also known as “scars” (e.g., chromosomal instability and other genomic signatures) [52]. Genomic scars are defined as aberrations that cause structural changes in chromosomes, such as loss of heterozygosity (LOH), telomeric allelic imbalances (TAI), and large-scale transitions (LST) [52]. Together, the analysis of these three aberrations—LOH, TAI, and LST—allows the calculation of a Genomic Instability Score (GIS), which can be used to define HRD status.

Depending on the level of analysis, HRD status can be classified into three categories. The first includes germline or somatic mutations in genes involved in the HRR pathway, identifying variants that contribute to HRD. The second is based on genomic scars or mutational profiles that reflect genomic instability, such as LOH, TAI, and LST. These alterations serve as indirect markers of HRD. The third category focuses on the functional status of the HRR system: the presence of functional HRR proteins or the measurement of DNA repair efficiency may provide insight into the functional capacity of the HRR system in tumor cells [54].

Several laboratory tests have been developed as in vitro diagnostics (IVDs) tools to detect HRD using Next-Generation Sequencing (NGS) data for the molecular analysis of specific genetic alterations. These tests facilitate the identification of HRD status in clinical settings for OC [60–62]. In routine diagnostic practice, these analyses are typically performed in pathology laboratories using validated NGS panels that focus on mutational and genomic analyses, while functional assays are not yet standardized or widely implemented [63].

HR status: clinical implications

A meta-analysis of 14 studies revealed a more favorable prognosis for women with OC bearing BRCA1/2 mutations [64]. These women demonstrated improved OS with a hazard ratio (HR) of 0.76 (95% confidence interval (CI) 0.70–0.83) for BRCA1 and an HR of 0.58 (95% CI 0.50–0.66) for BRCA2, as well as improved PFS with an HR of 0.65 (95% CI 0.52–0.81) for BRCA1 and an HR of 0.61 (95% CI 0.47–0.80) for BRCA2.

Germline and somatic BRCA1/2 mutations in patients with platinum-sensitive, recurrent OC have also been associated with greater sensitivity to PARP inhibitors in the maintenance setting, as demonstrated in the SOLO2 [65], NOVA [66], and ARIEL3 clinical trials [67]. Several randomized, phase III clinical trials (SOLO 1, PAOLA, PRIMA, PRIME, and ATHENA–MONO) have demonstrated that patients with high-grade serous or endometrioid OC tumors, classified as stage III or IV, who responded to initial platinum-based regimens experienced a significantly increased median PFS with maintenance therapy using PARPi (olaparib, niraparib, olaparib plus bevacizumab, and rucaparib) [68–72]. These trials consistently demonstrated the substantial benefits of olaparib–bevacizumab, niraparib, and rucaparib for patients with BRCA1/2 mutations and the HRD population.

Notably, niraparib and rucaparib showed benefits for both the HRP and HRD-unknown subgroups, though the benefits were smaller in magnitude. In contrast, adding olaparib to bevacizumab showed no benefit for patients classified as HRD-negative (HR 1.19; 95% CI 0.88–1.63) [73].

Recommendations for HRD testing

Pathologists play a pivotal role in accurate HRD testing in OC, ensuring proper sample collection, fixation, and tissue processing to maintain DNA integrity, as well as implementing validated genomic platforms for HRD assessment within their laboratories.

The preferred material for diagnosis and molecular analysis is a surgical specimen or image-guided biopsy from an untreated tumor. Pathologists should select samples containing sufficient tumor cells (ideally > 30%) while avoiding necrotic or highly inflamed tissue. When tumor tissue is unavailable, cell blocks from peritoneal or pleural effusions may be used if cellularity is adequate.

Multiple HRD assays are currently accessible for use in diagnostic laboratories equipped with NGS platforms. These assays integrate sequencing for the detection of mutations in HRR genes with the evaluation of structural genomic alterations indicative of HRD [74, 75]. Most of these assays also report the somatic mutational status of BRCA1/2. A detailed description or comparison of these assays falls beyond the scope of this article.

These consensus guidelines recommend assessment of HRD status at diagnosis for all patients with high-grade non-mucinous ovarian carcinoma, in FIGO stages III and IV. HRD testing provides valuable predictive information for selecting first-line maintenance therapy.

Tissue biomarkers

TP53/p53

The protein p53 is critical for maintaining genome integrity by coordinating multiple DNA damage response mechanisms, including DNA damage response repair, which leads to cell cycle arrest or apoptosis when damage is irreparable. Although p53 is not directly part of the HRR pathway, wild-type p53 helps activate the expression of HRR repair genes, like BRCA1 and BRCA2, thereby aiding in the repair of double-strand breaks (DSBs). Mutations in TP53, found in over 50% of human cancers, are particularly prevalent in HGSOC where they have been reported in 96% of cases [76]. These mutations occur in early stages of the disease and are considered drivers of ovarian carcinogenesis. Mutated p53 may lose its ability to regulate the HRR pathway, thus contributing to genomic instability and resistance to DNA damage repair, particularly to agents that induce DNA double-strand breaks (e.g., platinum-based therapies) [77]. The TP53 mutation status can be routinely assessed by IHC to identify abnormal p53 protein staining patterns, which are defined as either null or greater than 70% expression.

Folate receptor (FRα)

Folate is essential for nucleotide synthesis and DNA replication. It enters cells via the reduced folate carrier or folate receptor alpha (FRα), encoded by FOLR1, which has high affinity for folic acid. FRα supports DNA/RNA synthesis, cell growth, and proliferation and is widely expressed in ovarian cancer, where its overexpression correlates with poor prognosis and reduced chemotherapy efficacy [78]. FRα is evaluated using immunohistochemistry, although no standardized scoring system has been established.

The VENTANA FOLR1 (FOLR1-2.1) RxDx Assay is the only FDA- and EMA-approved IHC test to identify OC patients eligible for mirvetuximab soravtansine (MIRV), an FRα targeted antibody–drug conjugate (ADC). The clinical cut-off is ≥ 75% of viable tumor cells demonstrating moderate- or strong membrane staining (+ +/+  + +)[79].

MIRV (6 mg/kg IV) has shown improved PFS and OS in platinum-ineligible patients with high FRα expression and is currently being evaluated in platinum-sensitive settings, as monotherapy, combined with carboplatin, or as maintenance with bevacizumab among other settings [80, 81].

Several ADCs under development directed to FRα are being tested in different clinical trials that are also studying the value and relationship to FRα [82] (Table 2).

Table 2.

Emerging biomarker-targeted therapies in ongoing clinical trials for ovarian cancer

Drug	Target	Trial	Clinical phase	Results
Mirvetuximab Sorvtensine	FR alpha	MIRASOL	III	PFS 5.62 vs 3.98 months OS 16.46 vs 12.75 months
Farletuzumab	FR alpha	NCT03386942	I	ORR 25.0% and 52.4%
Luveltamab tazevibulin	FR alpha	STRO-002-GM1	Ib	ORR 37.5% PFS 6.1 months
Trastuzumab–deruxtecan	HER2	Destiny-PanTumor02	Ib	ORR 45.0% (HER2 3 + 63.6%) PFS 5.9 months (HER2 3 + 12.5 m) OS 13.2 months (HER2 3 + 20.0 m)
Nivolumab	MMRd/MSI-H	NCT03241745	II	ORR 58.8% PFS 21.6 months OS not reached
Pembrolizumab	MMRd/MSI-H	KEYNOTE-158	II	ORR 33.3% PFS 3.5 months
Adavosertib	CCNE1 amplification	NCT03253679	II	ORR 36% PFS 6.3 months
	CCEN1 overexpression	IGNITE	II	ORR 53%
Raludotatug–deruxtecan	Cadherin-6	NCT04707248	I	ORR 46% PFS 7.9 months
Datopotamab–deruxtecan	TROP2	TROPION-PanTumor03	II	ORR 42.9% PFS 5.6 months
Sacituzumab–tirumotecan	TROP2	MK-2870–001	I/II	ORR 40%mPFS 6.0 mo

HER2

HER2, encoded by the ERBB2 oncogene, is located on chromosome 17q12. HER2 testing is the standard of care in breast cancer, significantly improving outcomes [83]. In OC, studies remain limited and often small (n < 20 patients), reporting overexpression rates ranging from 0 to 45.6%. The evaluation method follows the 2018 ASCO IHC Guidelines for breast cancer, with FISH required for IHC 2 + cases [84]. Emerging DNA/RNA NGS approaches require further validation.

Trastuzumab-deruxtecan (T-DXd), a HER2-targeted ADC, showed efficacy in the phase II DESTINY-PanTumor02 trial across HER2-expressing solid tumors, including OC with an ORR of 45.0% (95% CI 29.3–61.5); in HER2 IHC 3 + OC, the ORR increased to 63.6% (95% CI 30.8–89.1). FDA approval was granted in April 2024 for HER2-positive (IHC 3 +) solid tumors lacking therapeutic alternatives [85].

Although anti-HER2 therapy is not standard in OC, HER2 testing is recommended at diagnosis in all high-grade tumors to facilitate clinical trials access.

Mismatch repair status

The MMR pathway corrects DNA replication errors [86]. MMR deficiency (MMRd), microsatellite instability-high (MSI-H) status, and germline MMR gene mutations occur most frequently in endometrioid and clear cell carcinomas, and least in serous tumors [87]. IHC for MLH1, MSH2, MSH6, and PMS2 is recommended; ambiguous cases benefit from MSI testing or MLH1 promoter methylation analysis. Some HRD-focused NGS panels also report MSI and MMR alterations.

In the KEYNOTE-158 trial, pembrolizumab demonstrated an ORR of 34.3% globally in MSI-H/MMRd tumors and 33.3% among 15 OC patients [88], leading to full FDA approval for MSI-H/MMRd solid tumors progressing after prior therapy.

A phase II study of nivolumab, in MMRd/MSI-H endometrial and ovarian cancer (35 patients), showed an ORR of 58.8%, with median PFS of 21.6 months and median OS not reached [89].

CCNE1 amplification

Amplification of CCNE1 (Cyclin E1) drives the G1/S transition and promotes genomic instability. It occurs in ~ 19% of OC, predominantly in high-grade serous carcinoma (HGSOC) [7, 90], correlating with tumor progression and poor prognosis [91]. Assessment is made via IHC (H-score > 50) and FISH (≥ 8 copies) [92]; alternative IHC + CISH criteria (≥ 60% positive cells, ≥ 5% strong staining) show good sensitivity and specificity [93].

A high level of CCNE1 amplification is mutually exclusive with BRCA1/2 germline mutations and may predict poor response to PARP inhibitors (a negative predictive biomarker) [93, 94].

Adavosertib, a WEE1 inhibitor, produced an ORR of 36% (median PFS 6.3 months) in CCNE1-amplified tumors [95] and an ORR of 53% in HGSOC with CCNE1 overexpression without amplification [92].

Cadherin-6

Cadherin-6 (CDH6) is a transmembrane protein involved in cell-to-cell adhesion, organ development and the epithelial-mesenchymal transition [96]. It is overexpressed in OC, and is associated with poor prognosis and lymph node metastasis [97]. IHC scoring often follows HER2 gastric guidelines (0–3 +), with ≥ 1 + considered positive. CDH6 expression is common across primary and recurrent OC, especially in HGSOC [97].

The ADC Raludotatug-deruxtecan (R-Dxd) demonstrated acceptable tolerability and promising activity in a phase I trial with an ORR of 46%, median DOR of 11.2 months, and median PFS of 7.9 months in heavily pretreated patients, irrespective of CDH6 levels [98]. A phase II dose optimization study is ongoing.

TROP-2

Trophoblast cell surface antigen 2 (TROP-2) is a calcium signal transducer implicated in proliferation and drug resistance [99]. IHC scoring combines percentage of membranous staining positive cells (0–4) and intensity (0–3), producing a total score (multiplying the two components) ranging from 0 to 12. TROP-2 overexpression is defined as > 4 [100, 101]. Strong expression is common in serous and endometrioid OC, moderate in mucinous OC, and weak/negative in clear cell histology [102]. Approximately 50% of OCs overexpress TROP2 [100, 101].

Datopotamab deruxtecan (Dato-DXd), an ADC targeting TROP-2, achieved an ORR of 42.9%, DCR of 91.4%, and median PFS of 5.8 months in the OC cohort of TROPION-PanTumor03 [102].

DB-1305 (BNT325) showed a DCR of 82.8% and a median DoR of 7.3 months [103].

Sacituzumab tirumotecan (Sat-TMT) reported an ORR of 40%, DCR of 75%, median DoR of 5.3 months, and PFS of 6.0 months in heavily pretreated OC patients [104].

ARID1A

ARID1A is a tumor-suppressor gene encoding a subunit of the SWI/SNF chromatin-remodeling complex. Loss-of-function mutations are common in endometrioid and clear cell ovarian carcinomas and are associated with genomic instability, impaired DNA repair, and unique vulnerabilities. ARID1A-deficient tumors show increased immune infiltration and may be more responsive to immunotherapy.

IHC for ARID1A (BAF250a) is widely used, with complete loss indicating a pathogenic mutation. ARID1A alterations frequently co-occur with MMR deficiency, reinforcing the importance of MSI/MMR testing in these histotypes.

Therapeutic strategies under study for ARID1A-mutated OC include EZH2 inhibitors, ATR inhibitors, and PARP inhibitors in specific synthetic-lethal contexts. ARID1A status is increasingly viewed as a potential predictive biomarker, particularly in endometrioid and clear cell OC [105] (Table 3).

Table 3.

Recommendations for biomarker testing in ovarian cancer

Biomarker category	Specific test	Recommendation level	Clinical application
Histopathological classification
General approach	Histologic examination + selective IHC/molecular testing	Mandatory		Diagnostic classification; use ancillary techniques for challenging or overlapping morphologies
High-grade serous carcinoma
General approach	p53, WT1, PAX8, ER (IHC)	Optional		Diagnostic confirmation when morphology is non-classic
	HRD testing	Mandatory		At diagnosis for stages III–IV to guide PARP inhibitor therapy
Endometrioid carcinoma
General approach	ER, PR, PAX8, CK7, WT1 (IHC)	Optional		Diagnostic confirmation when needed
	Molecular subgrouping (POLE, MMR, TP53, NSMP)	Optional		Prognostic stratification and treatment selection
	HRD testing	Mandatory		At diagnosis, if high grade, for stages III–IV
Clear cell carcinoma
General approach	Napsin A, HNF-1β, PAX8 (IHC)	Optional		Diagnostic confirmation
	ARID1A (IHC)	Optional		Loss supports diagnosis and identifies potential therapeutic targets
	MSI/MMR testing	Optional		Identifies immunotherapy candidates
	HRD testing	Mandatory		At diagnosis for stages III–IV
Low-grade serous carcinoma
General approach	WT1, ER, p53 (IHC)	Optional		Distinguish from HGSOC when needed
	MAPK pathway mutations (KRAS, BRAF, NRAS)	Investigational		May guide targeted therapy in clinical trials
Mucinous carcinoma
General approach	CK7, CK20, CDX2, PAX8, WT1 (IHC)	Optional		Differentiate primary from metastatic disease
	HER2 amplification	Optional		Prognostic and potential therapeutic value
	KRAS mutations	Optional		Prognostic and potential therapeutic value
Other rare carcinomas
General approach	Extended IHC panel (GATA3, TTF-1, CD10, MMR)	Optional		Subtype classification
	HRD testing	Mandatory		At diagnosis for stages III–IV
Serum biomarkers
CA-125	Serum measurement	Mandatory		Treatment monitoring and relapse detection (GCIG criteria); NOT for population screening
HE4	Serum measurement	Optional		Combined with CA-125 to improve diagnostic specificity, especially in premenopausal women
Multivariate indices (RMI, OVA1, ROMA)	Algorithmic tools	Optional		Differentiate benign from malignant adnexal masses when ultrasound is inconclusive
Treatment response
Chemotherapy Response Score (CRS)	Histologic assessment	Mandatory		Evaluate omental specimens post-neoadjuvant chemotherapy in HGSOC; CRS3 predicts improved PFS/OS
CA-125 elimination kinetics (KELIM)	Calculated parameter	Optional		Predictive marker for treatment response
Molecular biomarkers
HRD status	NGS-based assays	Mandatory		At diagnosis for all high-grade non-mucinous carcinomas, stages III–IV; requires ≥ 30% tumor cellularity; guides PARP inhibitor maintenance
BRCA1/2 (somatic)	NGS or targeted sequencing	Mandatory		Prognostic and predictive value; if positive, perform germline testing
BRCA1/2 (germline)	Genetic testing	Mandatory		When somatic mutation detected or family history suggests hereditary syndrome
TP53/p53	IHC	Mandatory		Confirm HGSOC diagnosis; molecular subgrouping in endometrioid carcinoma
Folate receptor-α (FRα)	IHC (validated assay)	Optional		Identify candidates for mirvetuximab soravtansine (requires ≥ 75% cells with moderate/strong staining)
HRR genes (RAD51C/D, PALB2, BRIP1, ATM)	NGS	Optional		If mutation detected, perform germline testing
HER2	IHC/FISH	Optional		In high-grade tumors for clinical trial eligibility or HER2-targeted therapy (e.g., trastuzumab deruxtecan)
MMR/MSI	IHC (MLH1, PMS2, MSH2, MSH6) + MLH1 methylation	Optional		Identifies immunotherapy candidates (MSI-H/MMRd); consider in clear cell and endometrioid carcinomas
CCNE1 amplification	IHC/FISH/NGS	Investigational		Poor PARP inhibitor response; may guide WEE1 inhibitor trials
ARID1A	IHC	Optional		Loss supports diagnosis in clear cell/endometrioid carcinomas; identifies potential therapeutic targets
Emerging biomarkers
Cadherin-6 (CDH6)	IHC	Investigational		For ADC trial eligibility (e.g., raludotatug deruxtecan); ≥ 1 + staining required
TROP-2	IHC	Investigational		Guide enrollment in ADC trials (Dato-DXd, DB-1305, Sat-TMT)
RAD51 foci	Functional assay	Investigational		Potential HRD surrogate; validation ongoing

Conclusions

The establishment of evidence-based guidelines for the use of biomarkers in ovarian cancer is essential for appropriate diagnosis, prognostic assessment, and personalized treatment strategies. Collaboration among medical teams involved in ovarian cancer diagnosis and treatment, and institutional support are crucial for achieving excellence in this field.

CA-125 remains a pivotal biomarker for monitoring treatment response and detecting recurrence. Tools like the KELIM score, which evaluates the kinetics of CA-125 decline during chemotherapy, provide valuable prognostic insights, particularly in assessing chemosensitivity and predicting patient outcomes.

Genomic biomarkers, such as BRCA1/2 mutations, play a central role in identifying patients who may benefit from PARP inhibitors, now a key treatment for the management of ovarian cancer. Furthermore, ongoing research into novel therapies like antibody–drug conjugates (ADCs) underscores the importance of integrating emerging biomarkers to predict and monitor responses to these innovative treatments.

This guideline aims to integrate both established and emerging biomarkers into clinical practice, offering practical recommendations while acknowledging the evolving landscape. Advances in biomarker discovery and validation will continue to enhance precision medicine and improve ovarian cancer patient outcomes.

Author contributions

All authors contributed to the study concept and design. All authors participated in the writing of the first draft of the manuscript, commented on subsequent draft versions and approved the final version of the manuscript.

Funding

This document was funded by the Spanish Society of Medical Oncology (SEOM) and the Spanish Society of Pathology (SEAP) with an unrestricted grant from ABBVIE, GSK and ASTRAZENECA.

Data availability

Not applicable

Declarations

Conflict of interest

The authors declare that, when writing and revising the text, they did not know the names of the pharmaceutical companies that provided financial support for this project, so this support has not influenced the content of this article.

Ethical approval

The study has been performed in accordance with the ethical standards of the Declaration of Helsinki and its later amendments. This article does not contain any studies with human participants or animals performed by any of the authors.