It is now well known that tumor cells lacking BRCA1/2 are highly sensitive to DNA damaging agents like platinum derivatives, as a consequence of impaired genomic damage repair which is induced by different mean 1923. Platinum compounds, through adduct formation at the DNA produce DBS, are specifically active in tumors with HR impairment for BRCA1/2 lack of function 24. There is strong evidence from preclinical and clinical studies for a specific sensitivity of different BRCA1/2-related tumors to platinum derivatives. In 2003 our group has demonstrated a differential chemosensitivity profile in vitro of BRCA1-mutated HCC1937 breast cancer cells with bi-allelic loss as compared to the derivative clone HCC1937^WT, where the BRCA1 expression has been reconstituted by transfection of a BRCA1 full length cDNA 25. This study led to the conclusion that HCC1937 are highly sensitive to Cisplatinum (CDDP) as compared to Estrogen Receptor expressing and BRCA1-competent MCF-7 and BRCA1-reconstituted HCC1937^WT. In the same work it was demonstrated that BRCA1-defective HCC1937 breast cancer cells were resistant to paclitaxel as compared to MCF-7 and HCC1937^WT cells. In a further study by our group a differential chemosensitivity of BRCA1-mutated HCC1937 human breast cancer cells to microtubule-interfering agents has been found 26 Quinn et al. in a similar cell system, where BRCA1 was reconstituted by a retroviral vector containing the full length cDNA and compared to parental cells infected with empty vector, demonstrated that BRCA1 functions as a differential modulator of chemotherapy-induced apoptosis. 27. In a recent follow up study 28, our group evaluated the in vivo differential chemosensitivity of BRCA1-defective versus BRCA1-reconstrituted xenografts in SCID mice. In this mouse model, we confirmed a differential and higher activity of CDDP against HCC1937 BRCA1-defective xenografts. Furthermore, we demonstrated a major difference in the whole gene expression profile by cDNA microarray. Specifically, we found reduced expression of ERCC1 and RRM1 in HCC1937 versus BRCA1-reconstitued parental cells. Importantly, these two genes have been demonstrated to correlate, in previous studies in lung cancer, to an impaired response to CDDP treatment and to improved survival in patients undergoing treatment with CDDP/gemcitabine 2930. In addition, we found increased expression of mRNAs for RAD52 and XRCC1, genes related to DNA damage recognition. These latter findings strongly suggest a compensatory response to the impaired DNA damage repair in BRCA1 defective cells.

All together these experimental observations suggest that sensitivity to platinum derivatives inversely correlates to sensitivity to paclitaxel in BRCA1-defective breast tumor cells. Quinn et al. 31 have recently demonstrated a direct correlation between BRCA1 mRNA expression levels and overall survival in patients with ovarian cancer undergoing chemotherapy. These authors have shown that inhibition of BRCA1 expression in ovarian cancer cell lines increases cell sensitivity to platinum derivatives, while reduces the antitumor activity of taxanes. Subsequently they have evaluated BRCA1 mRNA expression in 70 tissue samples in sporadic ovarian tumors from patients which underwent treatment with platinum derivatives and they found that patients with low-intermediate levels of BRCA1 mRNA had a significantly better outcome in terms of overall survival (OS) as compared to high BRCA1 mRNA levels (57.2 months versus 18.2 months p = 0.0017). Finally high BRCA1 mRNA tumor carriers had a better survival if treated with taxanes even if statistical significance was not reached. It was concluded that BRCA1 mRNA expression levels correlate in sporadic ovarian cancer patients with OS and can be considered a predictive marker of treatment response. More recently the same authors have produced a systematic review 32 on all published studies on this topic from 1990 to 2008 leading to the hypothesis of a possible role of BRCA1 as biomarker predictive of treatment response in hereditary and sporadic ovarian tumors. The authors conclude that the identification of a functional deficit in BRCA1 and in related pathways is likely to provide information on treatment efficacy. Finally, the same authors have provided evidence that BRCA1 protein expression may be a predictive marker of chemotherapy response in sporadic epithelial ovarian cancer. They found that BRCA1 protein expression is associated with a better outcome of platinum/taxane combination as compared with CDDP alone, while when BRCA1 protein was not detected CDDP alone was as effective as combination chemotherapy. These data indicate again that CDDP alone may be highly effective in the case of BRCA1 impairment 33.

It is a common finding that human tumors highly sensitive to chemotherapy may become resistant 34. Recent studies have shown that even the increased sensitivity to CDDP or Poly ADP Ribose Polymerase (PARP) inhibitors (see below) produced by BRCA1/2 gene mutation is not a stable trait 35. Stacey Edwards et al have demonstrated that when pancreatic cells with BRCA2 inactivation become resistant to PARP inhibitors, novel BRCA2 isoforms were detected in the resistant line resulting from intragenic deletion of the c.6174delT mutation and restoration of the open reading-frame (ORF) 36. Similarly Wataru Sakai et al. reported that secondary mutations in BRCA2 might reconstitute resistance to CDDP and PARP inhibitors in BRCA2 mutated tumors and that similar molecular mechanisms should be involved in clinical resistance to CDDP by ovarian tumors as demonstrated on clinical specimens 37.

All together these findings indicate that BRCA1/2 gene-mediated sensitivity to anticancer treatment can be reverted by escape mutations and that these important events must be taken in account for the design of novel therapeutic strategies in this specific setting.

Preliminary clinical evidence appears in line with preclinical in vitro findings and indicates that prospective clinical trials must be designed to clarify the clinical relevance of the differential sensitivity to anticancer drugs by BRCA1/2 mutated tumors.

In the last few years, the identification of individuals carriers of inactivating mutations on BRCA1 and 2 genes has been essentially directed to cancer prevention by the use of prophylactic surgery 3839, preventive treatments or screening procedures different from general population 40414243. Even with such measures, the onset of ovarian cancer in mutation carriers is a common event for failure of preventive strategies or because cancer predisposing mutations have been identified when the cancer has been already diagnosed. It is therefore often necessary the use of systemic chemotherapy regimens, which at present do not differ from those which are utilized in sporadic tumors. At present, ovary tumors are still a fatal disease in a high percentage of patients, due to late diagnosis for the lack of symptoms in early stage disease and partially for the intrinsic biologic aggressiveness. Systemic treatments are needed not only in advanced/metastatic disease but also in early cases and are based on the use of agents like platinum derivatives, taxanes, topotecan and liposomal doxorubicin.

The hereditary variants are characterized by typical clinical features, which often allow the selection of patients for genetic counseling and testing procedures, as early onset, multiple tumors especially in the breast, and family history for the same or other BRCA1/2 related tumors.

The increased risk of developing a ovary tumor in BRCA1/2 mutation carriers, assessed as 15,3% of the whole ovary cancer patients 44, has been related to the different functions of BRCA1 and 2 genes in the regulation of cell growth, genomic stability and repair of genomic damage by homologous recombination. Specifically, this last feature is the cause of the DSBs repair failure resulting in genomic instability and predisposition to neoplastic transformation for loss of function of BRCA1/2.

As regard to the clinical outcome of BRCA1/2-related breast tumors, several studies have been done in order to evaluate if germ-line cancer predisposing mutations might be useful for inclusion in different prognostic subgroups 43. The majority of published studies has not produced proof of prognostic value of BRCA1/2 inheritance. Such studies present however several majors flaws for the retrospective design and because patients where not considered on the basis of stage, age, histology and residual disease after primary surgery38.

A recent case-control study 45 included 779 jewish women affected by hereditary ovarian cancer who had undergone genetic testing for three Askhenazi founder mutations (BRCA1 185delAG, 5382insC; BRCA2 617delT). The design of the study was based on the comparison of mutation-positive versus mutation-negative ovarian cancer carriers in terms of long term outcome. The two groups were homogeneous for known prognostic, clinical and demographic factors.

This study clearly demonstrated a significantly better 5 years survival in mutation carriers as compared to non-carrier individuals (34.4% surviving in the non carriers versus 46% in the BRCA mutation carriers p = 0.003). The survival gain occurred in advanced stages but not in early stages and at multivariate analysis, the prognostic weight of BRCA1/2 mutation was independent from age at diagnosis, histology and grading. Subgroup analysis demonstrated a better outcome for BRCA2-related versus BRCA1-related or BRCA-unrelated, while BRCA1-related did not behave favorably if compared to the two other subgroups. Interpretation of these subgroup analysis needs caution however at this point and confirmation in larger studies is eagerly awaited.

Data from this study appear of interest but it has to be considered that it is difficult to generalize these findings to non-Askhenazi population with a more heterogeneous mutational status, to evaluate the impact of BRCA1 versus BRCA2 as well as to speculate on the potential role of other confounding factors or modifier genes which might by themselves retain a prognostic weight. Moreover this study doesn't allow to understand if the survival advantage achieved in BRCA1/2 mutation carriers as compared to non carriers might be related to intrinsic biologic features or to a better response to treatment. In any case this landmark study provides proof of principle that ovarian cancer arising in BRCA1/2 mutation carriers is a different disease.

Norah Kauff 3846 has discussed these important findings in a provocative Journal of Clinical Oncology editorial. The identification of BRCA1/2 related ovarian cancer as a distinct disease has important implications. Imbalance of BRCA1/2 related ovarian tumors in the arms of a randomized trial will introduce a powerful bias. It can be therefore inferred that all ovarian cancer patients enrolled in prospective randomized trials might be stratified on the basis of presence or absence of BRCA1/2 cancer predisposing mutations. These points merit further discussion. BRCA1/2 testing is an expensive procedure and has important ethical/consent implications. We think that prospective genetic testing cannot be performed in unselected individuals. We think however that enrollment in ovarian cancer clinical trials should be reserved to clinical centers offering genetic counseling to all ovarian cancer patients. Genetic testing based on pretesting counseling will allow the identification of most BRCA1/2 related tumors. In any way, Kauff's point needs to be considered in the planning of future clinical research in ovarian cancer.