Molecular and clinical aspects of borderline ovarian tumours

Abstract

In this thesis the molecular and clinical aspects of borderline ovarian tumours are discussed. In chapter 2 the prognostic and clinical value of morphometry and DNA cytology is studied in 93 borderline ovarian tumours (BOTs) after previous research that claimed prognostic power for these indicators. It is concluded that morphometry and DNA cytology are not useful in direct clinical management of BOTs, while histologic subtype and FIGO stage are stable prognostic indicators. In chapter 3 a report is presented of three cases of serous borderline ovarian tumour (SBT) with supradiaphragmatic lymph node involvement. In all 3 cases the first symptoms included cervical lympadenomas that after fine needle aspiration showed adenocarcinoma of unknown primary (ACUP). Subsequent extensive examination revealed a SBT in all 3 cases that could be treated curatively by surgery. It is concluded that extra abdominal spread of implants accompanying an SBT is possible and when ACUP is found and SBT should be in the differential diagnosis. In chapter 4 mutations in the BRAF gene were determined in 30 SBTs and their accompanying implants and the effect on clinical behaviour is established. A BRAF V599E mutation was found in 41% of SBTs. BRAF mutation positive SBTs were frequently bilateral (36%) tend to present with lower FIGO stage, higher tumourvolume and are less frequently aneuploid. We hypothesize that BRAF mutation positive SBTs might form a subgroup with a relatively benign clinically behaviour. In chapter 5 we describe array-comparative genomic hybridization of DNA material from tumour cells of a progressive SBT in time. It is concluded that even progressive SBTs do not seem to harbor many genomic imbalances and based on this study we conclude that gains at 1q31-41 and 12q12-24 and losses at 1p12-36 may play a role in clinical progression of SBTs. In chapter 6 the genomic differences between SBTs and high grade serous ovarian carcinomas (OVCAs) were studied by multiplex ligation dependent probe amplification (MLPA). We found that SBTs harbour much less genomic imbalances compared to serous OVCAs. From the dataset (18 SBTs and 20 OVCAs) six mathematic models were derived to predict whether tumour cell DNA is derived from an SBT or an OVCA. These models were then tested on a second set of 23 SBTs and 27 OVCAs. We conclude that these results subscribe the theory that SBTs are no precursor lesions of high-grade serous ovarian carcinomas. In chapter 7 a case of LOH in the BRCA2 gene is described in a patients with a SBT. This patient was a known germ-line BRCA2 mutation carrier and developed an SBT. In order to investigate whether this SBT could be attributed to the BRCA2 mutation status and therefore could be added to the known BRCA2-related tumour spectrum, tumour DNA was tested to find if the wild-type BRCA2 gene was lost. Instead we found loss of the mutated BRCA2 gene. In addition a literature search was conducted for a correlation between SBTs and BRCA 1/2 mutations which could not be confirmed. It is concluded that SBTs are not part of the BRCA 1/2 related tumour spectrum. In chapter 8 it is concluded that SBTs are not likely to be precursors of high grade serous ovarian carcinoma but can rarely progress via an entity called micropapillary serous carcinoma (MPSC) to low grade serous ovarian carcinoma. Furthermore, morphometry and DNA cytology are not good prognostic indicators in BOT, BRAF mutations are commonly seen in bilateral SBTs and their accompanying implants and may constitute a prognostic favourable subgroup, SBTs can present with extra-abdominal implants and SBTs are not part of the BRCA 1/2 related tumour spectrum.