Ovarian cancer (OC) is diagnosed in a large number of women around the world annually - in 2020, 313,959 new cases were reported - significantly reducing their comfort level and also taking the lives of many of them (207,252 deaths) (1, 2). OC ranks seventh in cancer incidence and is fifth leading cause of cancer-related death in women worldwide (1), in addition is considered to be the most lethal gynecological cancer (3). Statistical studies have shown a general downward trend in the incidence and mortality of ovarian cancer over the past decade linked, no doubt, to advances in medicine. However, the significant growth in incidence in younger women is worrisome (2).

It is known that the increased risk of developing ovarian cancer is related to genetic predisposition and family history (Hereditary Ovarian Cancer - HOC). Hereditary cancers, i.e. those caused by inherited mutations in genes that predispose to cancer development, account for as many as 20-25% of epithelial ovarian cancer cases (4). It is worth mentioning that breast and ovarian cancers show some common features at the molecular level (5). The distinct disease syndrome is Hereditary Breast and Ovarian Cancer (HBOC) in which the risk of breast and ovarian cancer is higher than in the general population (6). Breast cancer cases due to HBOC are estimated to account for 2 to 7%; for ovarian cancer, the range is 15 to 18% (4).

Whether spontaneously occurring or associated with a genetic predisposition, OC represents a major global problem and a challenge of modern medicine. Undoubtedly, the development of next-generation sequencing (NGS) technology has contributed to a better knowledge and understanding of cancer biology and, moreover, in the detection of high-risk groups (7). This is crucial in the early diagnosis of cancer and also in the selection of targeted therapy (8). NGS panels approved for use in patient diagnosis allow assessment of both somatic and germline variants, as well as single or global genomic changes (7). Such broad capabilities of NGS technology make it widely used in scientific research and patient diagnostics, including oncology, and an important step towards personalized medicine. However, many factors influence the success and improvement of NGS research, such as sequencing chemistry, sequencing technology, bioinformatics and data analysis (9).

High-grade serous ovarian cancer (HGSOC) is the most common histological type of ovarian cancer and accounts for about 75% of ovarian cancers of epithelial origin (10). Despite the common histological features identified by pathologists, HGSOC tumors have significant differences in the clinical course of the disease, including response to treatment (11). HGSOCs exhibit significant heterogeneity due to differences at the molecular level (12).

The TP53 somatic mutation occurs as the first molecular alteration involved in HGSOC carcinogenesis and is referred to as a driver mutation. This condition was identified in 96% of HGSOC cases (3). The protein encoded by TP53, called p53, is a transcription factor that activates genes engaged in DNA repair, cell cycle and apoptosis after nonreversible DNA damage (13). Loss of p53 function promotes tumorigenesis, but does not itself cause a malignant phenotype, and at least one more genotoxic event is required. Most frequently, this is a BRCA1/2 mutation. BRCA1/2 can lose its functionality through somatic mutation, germline mutation or through promoter hypermethylation, which belongs to epigenetic modifications. It is worth noting that mutations are mutually exclusive with epigenetic silencing of the BRCA1/2 gene (3). BRCA1 and BRCA2 proteins are involved in DNA repair through homologous recombination (homologous recombination repair (HRR)) (14). Defects in HRR are estimated to occur in 50% of HGSOC cases. Besides changes in BRCA1/2, it can be caused by mutations in ATM, BRIP1, CHEK2, NBN, PALB2, RAD51B ( 3, 14). However, an HGSOC tumor can acquire genomic instability through disruption in another pathway. BRCA1/2 loss of function is mutually exclusive with CCNE1 amplification and RB1 inactivation. Both CCNE1 and RB1 are involved in the Rb cell cycle regulation pathway (3). The protein cyclin E1 forms a complex with cyclin-dependent kinase 2 (CDK2) to promote cell cycle progression from G1 to S phase (15). Cyclin E1 expression relies on E2F transcription factors that are bound to retinoblastoma protein (Rb) (16). Amplification of CCNE1 and deletion of RB1 hasten the cell cycle, leading to defective S-phase progression and an elevated number of chromosome breaks, resulting in genomic instability. CCNE1 amplification appears in about 20% of HGSOC cases while RB1 inactivation occurs in 10% of HGSOC (3).

Currently, the standard of care for patients with stage II-IV ovarian cancer is platinum-based chemotherapy and surgical cytoreduction (16). Ovarian tumors with HRD are characterized by higher sensitivity to platinum and poly(ADP-ribose) polymerase inhibitor (PARPi) therapies. When RB1 function is lost, the tumor is also sensitive to platinum based therapies. In contrast, ovarian cancer with CCNE1 overexpression shows greater resistance to platinum compound therapies (17). Identifying the pathogenesis of the tumor at the molecular level allows the choice of targeted therapy.

Most cases of HOC and HBOC are determined by germline mutations in BRCA1 or BRCA2 ( 6, 42). It is estimated that about 1 in 400-800 individuals carry the germline pathogenic variant of the BRCA1 and/or BRCA2 genes (43). Patients are placed under medical supervision if diagnosed as carrying a BRCA mutation. Chemoprevention or risk-reducing surgical treatment: risk‐reducing mastectomy and risk‐reduction salpingo‐oophorectomy may be used to reduce the risk of developing breast and ovarian cancer (44).

As for familial or hereditary ovarian cancer syndromes, they are caused by inherited mutations in tumor suppressor genes. In addition to BRCA1 and BRCA2 genes belonging to the Fanconi anemia pathway, mutations in PALB2, ATM, RAD51C/D and BRIP1 are also known to increase the risk of ovarian cancer. Another important pathway is the mismatched DNA repair pathway and genes such as MLH1, MSH2 and MSH6 ( 42).

Beyond the increased risk of developing breast and ovarian cancer, individuals in the HBOC group are at increased risk of developing other cancers, such as melanoma, pancreatic cancer and, in the case of men, prostate cancer (6). The risk of developing a specific cancer is related to the type of mutation. Women who have inherited the pathogenic BRCA1 variant tend to suffer from ovarian cancer at a younger age. In contrast, patients with a mutation in BRCA2 have a greater predisposition to develop male breast cancer, prostate cancer, pancreatic cancer and melanoma (45). Other HBOC-related genes include ATM, BRIP1, CDH1, PALB2, PTEN, RAD51C, and TP53 ( 46). Velázquez et al. used a panel of 35 genes to study hereditary cancers. In the case of HBOC, pathogenic variants were noted in the following genes: BLM, BRCA2, BRIP1, CHEK2. The researchers underline that by using the expanded panel, they were able to detect pathogenic variants that would not have been detected with routinely used diagnostic panels (47). Resch et al. also demonstrated the benefits of using wider gene panels in their study. Most of the pathogenic variants identified were BRCA1, BRCA2. However, they also detected variants in CHEK2 and RAD51C and also in ATM, BARD1, MUTYCH and SMARCA4 which would not have been possible with a standard panel (48).

It should also be taken into account that the etiology of HBOC may be related to constitutional mosaic variants. These variants can involve somatic and also germinal cells, and can therefore be passed onto offspring. In their work, Hidalgo Mayoral et al. described a clinical case of a breast cancer patient in whom they detected a mosaic variant c.9648 + 1G>A in the BRCA2 gene. The authors point to other cases reported in the literature and emphasize the role of BRCA1/2 mosaic variants in HBOC. Due to the low allele frequency (VAF), many of these may have been missed in genetic analyses. A separate diagnostic algorithm is needed to detect potential mosaic variants using NGS sequencing (49).

Despite the significant genomic similarities between ovarian and breast tumors and the existence of the HBOC disease entity, these tumors show some distinguishing features. Weber-Lassalle et al. subjected samples of HBOC patients to genetic analysis. The researchers identified germline variants in BARD1 in breast cancer patients. Moreover, the presence of these variants was associated with an earlier onset of cancer (42.3 years) compared to the overall study group (48.6 years). BARD1 variants were not detected in ovarian cancer (50).

It is extremely important that testing is not limited to BRCA1/2. The benefit of expanding molecular analyses was demonstrated by Tao et al. 25% of HGSOC patients had genetic alterations other than disorders leading to BRCA loss of function. This gives patients a chance to receive off-label treatment or other therapies described in this review as part of their participation in clinical trials (51).

Currently available tests to assess HRD include: analysis of genes associated with homologous recombination repair, genomic “scars” and/or mutational signatures, and real-time functional assessment of HRD. It should also be remembered that the HRD phenotype is very complex and thus extremely difficult to assess reliably and unequivocally in a single analysis. These validated tests can be used to make treatment decisions for PARPi. However, they are not effective in predicting the group of patients who will not benefit from PARPi, i.e. they have no negative predictive value (23, 52).

Considering the mutually exclusive nature of the occurrence of BRCA mutations and CNNE1 amplification, the detection of high-level amplification in combination with CCNE1 overexpression can be used as a criterion for predicting PARPi resistance (53). However, there are reports that genetic changes primarily thought to be mutually exclusive can co-occur. This is particularly true for loss of RB function, which appears to co-occur in tumors with BRCA1/2 mutations or other disorders of homologous recombination pathway genes (54). Perhaps understanding this aspect will become possible through the use of spatial genomic, transcriptomic and proteomic technology. Spatial biology offers the possibility to analyze different cellular subclones and also complex interactions with the tumor microenvironment (TME), which together determine the process of carcinogenesis in an individual (55).

It should be noted that CCNE1 overexpression can occur without copy number amplification. Clinical trials have demonstrated the efficacy of WEE1i in both patients with CCNE1 amplification and overexpression. Therefore, future studies will probably evaluate the possibility of using CCNE1 expression instead of amplification of this gene as a marker of susceptibility to WEE1i therapy (56).

It is also worth mentioning that the use of PARPi prolonged PFS, but not OS. Therefore, it is important to consider the use of another therapeutic option. Unfortunately, immunotherapies in ovarian cancer have been less successful than in non-small cell lung cancer and melanoma. In order to improve the efficacy of this therapy, patient selection based on immunological profiling including assessment of high-grade microsatellite instability (MSIH), mismatch repair deficiencies (dMMR) and HRD is proposed (57).

In order for a patient to qualify for genetic testing, a woman must meet criteria outlined in national guidelines. These include age, hormone receptor status in BC, OC histology, and history of cancers that may be related to HBOC in the family. Although, these guidelines in different countries cover the same aspects however, the cutoff points may vary. There is a need to refine consistent recommendations between countries for monitoring mutation carriers specifically in genes other than BRCA1/2 ( 58).

The decision to use a multigene panel is up to the clinician. The physician, after analyzing the patient’s phenotype and taking a health history, can choose the most appropriate molecular diagnostic option to identify HBOC, or to exclude it with higher probability. In addition to commercially available panel tests with a wide range, some laboratories also offer custom-designed panels after listening to the clinician’s suggestions (45).

Despite improvements in diagnosis through the use of multigene panels, the issue of variants of uncertain significance (VUS) remains crucial. There is a need for algorithms as well as improved data flow and monitoring the situation globally to identify variants on which further research should be conducted to determine their clinical significance (59). These issues are discussed in more detail in the next section of the article.

Another important aspect is that HBOC tests should be tailored to different ethnic groups, as each has characteristic variants that occur with increased frequency compared to other groups. This approach can contribute to a better understanding of the molecular characteristics of specific communities of people and lead to the development of genetic tests dedicated to a specific community which will markedly improve the sensitivity and specificity of the test (60).

With the advent of sequencing technologies, we have seen their continued evolution. Sanger sequencing, referred to as first-generation sequencing, is limited to regions of interest within the genes of interest. The development of next-generation sequencing technologies has enabled comprehensive sequencing of larger regions of the genome. In the case of gene panels, gene fragments whose genetic variation is responsible for the occurrence of a defined phenotype are sequenced. It is also possible to sequence targeted protein-coding regions of all genes - whole exome sequencing (WES) and even whole genome sequencing (WGS), i.e. coding and non-coding sequences. NGS allows for more extensive analyses and is consequently more efficient due to the time and cost savings compared to Sanger sequencing (61).

Using advances in technology such as NGS, it is possible to make treatment decisions (62). Currently, biopsy material archived in the form of formalin-fixed paraffin-embedded (FFPE) blocks is used for molecular diagnosis of solid tumors (63). In this material, due to the fixation process, there may be changes in nucleic acids, their degradation and also modification of nitrogenous bases, which will be misread as a sequence variant in NGS analysis. To prevent these errors, researchers are developing new and more efficient protocols for nucleic acid isolation and library preparation (64). Another aspect of crucial importance in the fight against ovarian cancer worldwide is early detection and also monitor disease progression based on molecular testing. Many studies are looking into the possibility of using liquid biopsy, which has the undoubted advantage of being non-invasive (62). Analysis of cell-free tumor DNA (ctDNA) circulating in the blood, which can be performed using the NGS technique, has many potential applications such as early detection of both primary disease and recurrence, predicting prognosis, characterization of tumor heterogeneity and monitoring of treatment over time (7, 65). However, the researchers point out that protocols for analyzing material obtained by liquid biopsy need to be refined, ctDNA enrichment methods need to be improved and the sensitivity of analytical techniques for detecting at low allele frequencies needs to be optimized. Further research is needed before liquid biopsy can be used in clinical practice (66).

The discovery of mutually exclusive pathways and also their interactions has become the basis for the development of new targeted therapies and, in the future, diagnostic regimens that qualify patients for treatment (67). Also in the case of approved drugs - PARPi, whose action is based on the concept of synthetic lethality, incoming data from studies show that it is essential to know the genetic context of the pathogenic change (68). For better insight and understanding of the clinical course of the disease, it is necessary to take into account not only individual genomic changes but their complex interrelationships, which together make up the characteristics of a given patient and thus determine the success of the planned therapeutic strategy (69). Consequently, there is a need to develop new biomarkers, the combined use of which will enable a better understanding of the patient’s clinical condition and allow the selection of the optimal treatment pathway (68).

NGS sequencing has provided a wealth of data regarding the molecular biology of cancers, including HGSOC. However, it is now known that the use of bioinformatics analysis and tools is essential. It is through these that the pathways of carcinogenesis can be analyzed in depth (67). To detect homologous recombination deficiency, research uses methods based on whole-exome or whole-genome sequencing. However, in the practice of molecular diagnostic oncology, targeted sequencing panels are most often used. Therefore, the number of identifiable mutations is too low for HRD signature analysis, and as a result, many cases with HRD are not identified and patients do not receive PARPi treatment. Gulhan et al. propose a computational tool called Signature Multivariate Analysis (SigMA), which enables detection of mutation signatures from panel sequencing data. The researchers’ impressive results indicate that the use of SigMA could significantly increase the number of patients who could be considered for targeted therapy (70). Schouten et al. developed a classifier that detects HRD-associated mutations based on evaluation of sequence variants in a panel of genes, epigenetic changes, and copy number alteration. The classifier proved to be an effective tool identifying most cancers with BRCA loss-of-function and BRCA-like phenotype, as well as all analyzed tumors with the RAD51C germline mutation (71). In contrast, Li et al. created a signature predictive of initial platinum resistance (IPR) not dependent on HRD but based on the differential expression of 11 genes. The researchers also identified significant differences in the characteristics of tumor infiltrating lymphocytes. The results suggest that patients with a high-IPR signature should show a better response to ICB (72).

In order for knowledge of the molecular profiles of tumors to be realistically reflected in clinical practice, that is, for treatment to be applied to a patient, it is important to keep in mind the significant heterogeneity of OC tumors, both temporally and spatially. For this reason, it seems reasonable that sequencing should also be performed after treatment and in relapse (73). This involves more research and thus cost generation. In silico data can be used to validate NGS assays instead of physical samples. In addition to high throughput and reliability, the use of in silico data makes it possible not to generate incremental costs associated with sequencing additional samples. These data can be successfully used to simulate hard-to-find gene variants, low-frequency variants or specific classes of variants (e.g., medium-sized insertions) (74).

The clinical course of the disease varies depending on the gene in which the pathogenic variant is present, whether the variant is germline, somatic, or an epigenetic modification, and the location of the change, i.e. which codon is affected (75). Therefore, correct and clear reporting and classify of detected changes is crucial. Somatic sequence variants are classified based on their clinical significance: tier I - variants of strong clinical significance; tier II - variants of potential clinical significance; tier III - variants of unknown clinical significance (VUS); and tier IV - variants of known non-significant significance (considered benign or probably benign). According to current guidelines, variants from tiers I-III must be included on the NGS analysis report, while it is not recommended to include lesions classified as tier IV. Tiers I and II have therapeutic, prognostic and diagnostic significance. VUS variants are the most challenging, and it is essential to report them which will enable researchers to decide later on their reclassification (76). Whereas germline variants are classified in a five-tier system: class 1 – non-pathogenic; class 2 - likely not pathogenic; class 3 – uncertain; class 4 - likely pathogenic; and class 5 - definitely pathogenic (77). Due to the distinctiveness of the scale used in the classification of somatic and germline sequential variants, it is helpful to include a verbal category name that indicates the pathogenicity of the variant to avoid potential confusion and misunderstanding.

Also extremely useful in the classification of variants are informatics tools that allow for the efficient flow of incoming data from NGS analyses. Hirotsu et al. demonstrated the utility of using MH BRCA and MH Guide developed by Molecular Health in determining the significance of VUS. The authors show significant value in confirming or exclusion of HBOC and also in predicting susceptibility to PARPi application when a variant of unknown clinical significance is detected (78).

Modern medicine has a great knowledge of the molecular-level changes involved in carcinogenesis, resulting in breakthroughs in patient diagnosis and treatment. Approaching cancer as if it were a disease determined by genetic changes and using NGS technology can significantly improve the prognosis of women with HGSOC. The use and utility of NGS testing in OC is summarized in Figure 2 . However, despite its high efficiency, performance, NGS poses many challenges. Bioinformatics tools and the use of algorithms to obtain comprehensive and reliable molecular characterization of a patient’s sample are helpful. It is crucial to properly interpret the test results and relate them to the patient’s clinical condition. There is also a need to search for new therapeutic targets and conduct clinical trials of further targeted therapies. Before introducing new drugs into clinical practice, it is necessary to evaluate their efficacy, selectivity against tumor cells and also possible side effects. Clinical trials on the efficacy of new molecularly targeted drugs are being conducted around the world and Table 2 collects examples of such trials. A great opportunity may be the possibility of tumor-agnostic therapy, i.e. the use of targeted drugs regardless of the type of cancer just on the basis of finding the presence of a specific genetic change. Biomarkers should be sought to predict the response to the planned therapy. A very important aspect is the identification of women with a hereditary predisposition to ovarian cancer. As the range of therapeutic options grows, it is necessary to develop a laboratory diagnostic pathway that allows the best classification of patients for specific therapies. Pursuing these goals will contribute to more efficient diagnosis and treatment from an economic perspective and, invaluably, to improving the quality and prolonging the lives of patients. This can only be achieved through continuous improvement of skills and cooperation between clinicians and laboratory diagnosticians.

PB: Conceptualization, Data collection, Writing. JK: Supervision. JN: Supervision. All authors contributed to the article and approved the submitted version.