Breast cancer (BC) and ovarian cancer (OC) are the first and the eighth most common tumors for both incidence and mortality in females, respectively [1
]. BC can also affect males, even if male breast cancer (MBC) represents less than 1% of all BCs [2
As for other cancers, BC and OC are the result of a combination of both environmental and genetic risk factors. About 10–30% of BCs and OCs show familial clustering, but only 5–10% of cases are estimated to be hereditary, being associated with a germline pathogenic variant (PV) or likely-pathogenic variant (LPV) in a cancer predisposition gene [5
] and BRCA2
] are the main genes involved in Hereditary Breast and Ovarian Cancer syndrome (HBOC) [9
], but other genes have also been associated with BC and OC risk [10
In the last few years, the advent of Next-Generation Sequencing (NGS) has enabled the analysis of a great number of genes with the advantage of lower costs and wider access to molecular tests for patients with suspected genetic syndromes [18
In this complex scenario, one of the main issues is to define how many and which genes should be tested in patients with a suspicion of a genetic predisposition to cancer.
In this review, we summarize the past and most recent genetic findings on BC/OC predisposition, grouping the genes on the basis of their penetrance, calculated on large case-control studies, and taking into account their function and association with genetic disorders (Table 1
). The penetrance of a disease-causing genetic variant is the proportion of carriers of that variant who develop the disease, whereas the relative risk (RR) is the measure of the risk of developing a disease compared to the risk of the general population. A genetic variant is usually defined with high penetrance when the RR for the carrier is ≥10.0, with medium-high penetrance when the RR is between 5.0 and 10.0, with moderate penetrance when the RR is between 2.0 and 5.0, and with low penetrance when the RR is between 1.0 and 2.0 [19
4. Other Emerging Genes
Thanks to the wide use of NGS, the number of genes that have been associated with a BC/OC risk has hugely increased in the last few years. However, the most critical issue is the penetrance estimate of the penetrance of these PV/LPVs.
Recently, PV/LPVs in genes coding RecQ helicases have been reported in patients with BC/OC [117
]. Interestingly, PV/LPVs in three of these genes are responsible for autosomal recessive genetic syndromes, all associated with an increased risk of malignancies: In fact, BLM
, and RECL4
genes are the genetic cause of Bloom syndrome [255
], Werner syndrome [256
], and Rothmund–Thomson syndrome [257
The proteins encoded by the ABRAXAS1
genes (also known as FAM175A
) form a complex with BRCA1, required for DNA damage response [258
]. PV/LPVs in both genes have been linked to an increased risk of BC, but penetrance estimates are still missing [259
Other genes, such as PPM1D
, and NTHL1
have been reported in BC/OC families [117
], but further studies are needed to confirm these associations.
Moreover, in 2018, a systematic review of The Cancer Genome Atlas (TCGA) data was performed, focusing on pathogenic germline variants: besides the BRCA1
genes, PV/LPVs in ATM
, and APC
genes have been detected in BC patients, and PV/LPVs in ATM
, and SDHA
genes have been detected in OC patients [267
]. Some of these genes are intriguing candidates in the predisposition to BC/OC and, in the future, could be included in diagnostic panels for the assessment of cancer risk.
BC and OC are both in the top 10 of the most common and deadly tumors for women and, among risk factors for the development of these cancers, genetic predisposition plays an important role.
BRCA1 and BRCA2 have been known for decades to be predisposition genes to BC and OC. Consequently, for PV/LPV carriers in these genes, accurate cancer risk estimates are available, as well as surveillance protocols for cancer prevention and early detection of the disease.
New genes are constantly emerging from NGS studies [10
], showing that BC/OC predisposition is distributed over many genes, with only a few genes being recurrently mutated. However, testing on genes other than BRCA1
is not routinely performed, due to lack of information about the actual risks for PV/LPV carriers and the unavailability of surveillance programs.
, there are some genes with high penetrance, such as TP53
, and CDH1
, whose variants are associated with other cancer genetic syndromes (Table 1
) that, however, also include a high risk of BC and/or OC. Consequently, these genes need to be included in a panel of genes for the identification of patients at risk of BC/OC. Moreover, there are some other genes, such as PALB2
, and MMR genes that are classified as moderate or low penetrance genes for the risk of BC/OC (Table 1
), but the increasing evidence of associated cancer risks and the availability of recommendations for the management of the variant carriers suggest that they should be included, whenever possible, in a gene panel for BC/OC predisposition.
In particular, PALB2
is emerging as the most important BC predisposition gene after BRCA1
], but at the same time, it is still classified as a moderate penetrance gene. In the last few years, many studies pinpointed this gene as being associated with a high risk of BC and a high rate of bilateral BC [117
]. These observations further highlight both the high risk of BC associated with PALB2
PVs and the importance of adding the PALB2
gene to standard genetic tests for patients with suspected HBOC syndrome.
NGS-based approaches have also highlighted unexpected overlapping among genetic syndromes predisposing to BC/OC and gastrointestinal tumors, such as colorectal, gastric, and pancreatic cancers, raising the question of phenotypes associated with individual cancer susceptibility genes [273
]. These findings address the choice of wide panels, including the genes involved in the main cancer syndromes that can be used independently of the personal and family history of cancer. This creates new diagnostic opportunities but also increases the risk of an incorrect genetic diagnosis [275
On the whole, it is indisputable that NGS has deeply increased our knowledge of BC predisposition by increasing the number of susceptibility genes. However, because of the growing demand for higher throughput and lower costs, quality data and standardized procedures need to be carefully assessed. Moreover, genetic counseling and risk evaluation, as well as clinical management of patients and families at risk, are becoming more and more challenging [276
]. In particular, all health-care professionals who offer genetic testing must engage in constant education as the field is continuously evolving, with new data becoming available [277
]. An important aspect is the selection of patients for the genetic test, which is currently based on the number and type of cancers in the family and on the age of onset of these tumors. Several models are available to estimate the probability of having a BRCA1/2
]. These models can be used to select patients for a multigene panel test for the risk of BC/OC but, with the increase in the number of predisposing genes, they should be improved to identify individuals who can really benefit from a genetic test and, at the same time, to avoid the overuse of genetic tests.
Moreover, another important issue for patients with a cancer predisposition syndrome is prophylactic surgery. To date, we know that prophylactic surgery is an option that should be evaluated by patients with specific syndromes. In women with BRCA1/2
PV/LPVs, bilateral RRM and bilateral RRSO have been demonstrated as effective measures to prevent cancer [285
]. In individuals with LS, prophylactic hysterectomy and bilateral RRSO can be considered after childbearing is completed, whereas prophylactic colectomy before the development of colon cancer is generally not recommended because screening colonoscopy with polypectomy is an effective preventive measure [286
]. Finally, for individuals with CDH1
PV/LPVs, prophylactic gastrectomy should be strongly advised; however, there are insufficient data about prophylactic mastectomy [287
]. For individuals with PV/LPVs in other cancer genes, more data are needed to perform a risk-benefit assessment of prophylactic surgeries but, for some genes, guidelines suggest that RRM and/or RRSO should be evaluated in accordance with family history and, in particular, the age of onset of the tumors in the family [54
Recently, cancer genetic predisposition and precision medicine have found a contact point, thanks to the discovery of the therapeutic potential of PARP inhibitors in BRCA1/2
PV/LPV carriers [288
], at first in OC [290
] and then in BC, prostate, and pancreatic cancers [291
]. PARP inhibitors have shown their efficacy, not only in patients with germline and somatic BRCA1/2
], but also in patients with PV/LPVs in genes involved in the HR pathway with BRCA1 and BRCA2 proteins [294
], and clinical trials are currently addressing this [136
]. These results pave the way for the future use of PARP inhibitors in all tumors with a deficiency of the HR system, independently of the germline or somatic nature of the alteration.
These “personalized” treatments are possible thanks to the fact that tumors originating in individuals with a germline PV/LPV in a cancer predisposition gene usually have a specific “mutational signature” that reflects the pathway in which the gene is involved [298
]. Indeed, the main genes associated with BC/OC predisposition encode proteins that act as tumor suppressors being involved mostly in DNA damage repair processes, such as HR and MMR, and are strictly linked together. This aspect will become very important in the future because it can help in the identification of new genes involved in the predisposition and can guide the choice of the best therapeutic approach in terms of targeted therapies, chemoprevention, and prophylactic surgeries for a medicine personalized on the genetic characteristics of the patient.
The diagnostic use of multigene panels, instead of the traditional single-gene analysis, generates many advantages as well as some critical issues. Before the advent of NGS, turn-around times for genetic testing were long, in some cases more than 6–12 months, while nowadays new technologies provide results in less than a month in many cases. This short time is extremely useful for the affected individuals because the result of the genetic test can address or modify the surgical and therapeutic approach to the disease but, at the same time, it can generate issues in the genetic counseling and the management of the family, since the implications of cancer predisposition need time to be understood properly by the patient and by family members. The use of large gene panels can also create an increase in the number of false-positive findings [299
]. For this reason, a second confirmatory analysis is always recommended, possibly with another technique, such as Sanger sequencing [300
]. Another important issue is the interpretation of variants of uncertain significance (VUS), whose number increases exponentially with the increase of tested genes. Of note, the assessment of the pathogenicity of the genetic variants is based on guidelines that take into account many factors [301
], but this classification refers to the potential role in cancer development without taking into account the penetrance and the spectrum of the associated diseases. Many non-easily classifiable variants are identified by NGS studies and, although several techniques can now be used to investigate their pathogenicity [302
], efficient and accurate classification methods are still needed to translate theoretical information into clinical practice. The bioinformatics tools for the prediction of pathogenicity seem inadequate to classify many variants and to identify higher-risk patients [117
]. However, risk assessment of candidate variants is made difficult by the limited number of variant carriers and by the possible interference of different genetic and environmental factors. The multifactorial nature of BC/OC and the presence of predisposing variants in genes never included in panels for standard genetic tests are likely to further increase the complexity of the scenario.
Taking into account all these considerations, it is clear how difficult it is to find the right combination of genes to be tested in patients with a suspected genetic predisposition to BC/OC. At the international level, efforts are being made to achieve a consensus [303
], but the identification of a balance between costs for health systems and benefits for patients remains one of the biggest challenges for the future. Moreover, larger case-control studies are needed to better refine the penetrance estimates and to evaluate the correct preventive and therapeutic approach for each patient.