Short summary

High grade serous ovarian carcinoma is the most common and aggressive type of ovarian cancer. A better biomarker for ovarian cancer is needed, as CA-125, the biomarker used today, is neither very specific nor sensitive. In this study plasma from 20 patients with high grade serous ovarian carcinomas will be sequenced in order to identify structural mutations in circulating tumor DNA with and without knowledge of mutations found in the primary tumor. If the method tested is valid we might be one step closer to a better tool for early diagnosis, monitoring during follow-up and in tailored treatment. This could potentially increase survival rate and quality of life among this group of patients.

Rationale

Background:

Ovarian cancer is the second most common gynecologic cancer in Denmark with approximately 450 new patients every year. It is the most deadly gynecologic cancer and the 6^th most deadly cancer in Europe.

More than 90% of ovarian carcinomas are of epithelial origin. The high grade serous carcinomas (HGSC) comprise 70% of all epithelial ovarian tumors and accounts for the majority of deaths in relation to ovarian cancer. This is due to its aggressiveness and the fact that symptoms of ovarian cancer occur in advanced stages and are non-specific. About 70% of the cases are Stage III-IV at diagnosis. The 5-year survival rate for all stages is 30-40%, however stage I disease has a 5-year survival rate of about 85%. The majority of patients with advanced ovarian cancer relapse with a median time of 16 months.

CA-125 remains the best characterized biomarker in ovarian cancer. It is unfortunately not specific to ovarian cancer, since it can also be elevated in relation to other cancers and benign diseases. Furthermore, this marker is not elevated in a substantial portion of patients with advanced ovarian cancer. It is therefore neither sensitive nor specific enough to be used for screening of the general population. There is furthermore no evidence that intensive monitoring of CA-125 during follow-up improves overall survival rate or quality of life.

Cancers harbor somatic genetic alterations, which include single-base substitutions, insertions, deletions and translocations. These mutations occur at a negligible frequency in normal cells and therefore these mutations have the potential to be used as exquisitely specific biomarkers for detection of circulating tumor DNA (ctDNA). The mutations are patient specific and need to be identified for each patient. So far, benign tumors and non-neoplastic conditions have not been seen to give rise to ctDNA and a rise in ctDNA has not been detected in non-relapsing patients. Therefore, ctDNA seems promising as a patient specific follow-up marker after cancer treatment and the risk of over-diagnosis seems to be low.

Plasma from cancer patients contain cell-free tumor DNA released from cancer cells. The clearance is quite inefficient compared to normal cells and therefore a larger amount of DNA is released in to the blood. This DNA carries information about tumor burden and tumor mutations. Detection of mutations in TP53 with allelic frequencies of 2 to 65% have been found in plasma from patients with ovarian cancer along with a high level of ctDNA using deep sequencing. Mutant allelic frequencies in plasma reflected the clinical course of the disease better than CA-125. This method could work if it was optimized and thereby be able to detect ctDNA in plasma from patients with less advanced cancers.

Bettegowda et al found ctDNA detectable in >75% of patients with advanced stages of various types of cancer including ovarian cancer. The frequency of patients with detectable ctDNA increased with stage as well as the concentration of ctDNA. A decrease in survival rate was seen with increasing levels of ctDNA.

The usual way to confirm suspected ovarian cancer and cancer in general, is to take biopsies from the primary tumor either during surgery or ultrasonic guided before chemotherapy. Cancer is highly heterogeneous which means mutations might differ in different metastatic clones and within the primary tumor. One biopsy from a tumor may not be enough to detect all mutations and plan the most efficient therapy in the future. The solution to this might be detection of cancer mutations within plasma DNA, called liquid biopsy. Sensitive methods for detection of cancer using plasma DNA have the potential to be favorable in early detection, prognosis, tailored treatment, monitoring therapeutic response and detection of minimal residual disease.

The most common mutated gene found in relation to HGSC is TP53, which is mutated in 96-97% of the cases. At lower frequencies recurrent somatic mutations are also seen in the following genes; NF1, BRCA 1 and 2, RB1, CDK12, FAT3, CSMD3 and GABRA6, though in a smaller amount of cases than TP53. Furthermore, focal DNA copy number aberrations have also been detected in the following genes: CCNE1, MYC and MECOM.

A sensitive and specific test for HGSC is highly desirable. The aim should be detection of low volume disease which has a better cure rate. Development of biomarkers that are present in early carcinogenesis might be the key and make it possible to do early detection and tailored treatment, which is likely to increase the effectiveness of treatment and thereby the survival rate.

Aim:

The aim of this study is to assess whether detection of individually specific breakpoints in ctDNA in women suffering from HGSC is possible. If possible, it could be a more sensitive and specific way to detect and monitor recurrence in patients treated for HGSC. Today follow-up of patients treated for HGCS is very controversial because there is no evidence that it increases survival rate and quality of life when compared to no monitoring. That might be because the methods that are in use today are not sensitive and specific enough.

Therefore we want to see if we can answer the following questions that could be useful in monitoring or early detection of HGSC:

1. Monitoring: Whether or not can breakpoints be found in HGSC-tumors using next generation sequencing? Are breakpoints found in the tumors, detectable in ctDNA in plasma, from the same patients, by PCR-based methods?
2. Early detection: Is it possible, by next generation sequencing of ctDNA in plasma from HGSC patients, to detect genetic mutations or copy number aberrations, known to be mutated in this type of cancer, without prior knowledge of somatic mutations in the individual patient?

Methods and materials:

Breakpoints found in ovarian tumors will act as a personalized profile for tumor DNA and this can be found as ctDNA in plasma from the same person.

1. 20 patients with HGSC will be identified at Odense University hospital (OUH). Exclusion criteria is if they have received neoadjuvant chemotherapy or they are suffering from a synchronous cancer. From each HGSC removed as part of the primary surgery, a tumor sample will be taken for this project. The samples will undergo next generation sequencing of the entire genome with low coverage in order to find breakpoints that are specific for each ovarian tumor. This will be done using Illumina HiSeq 1500 on large DNA fragments with paired end sequencing 2x100 bases.
   Then mapping of the read will be done using GATK pipeline including NOVOALIGN
   Blood leukocytes will be used as reference germline DNA. Copy number analysis will be performed using Nexus software
   A blood sample of 20 ml will be taken from each patient. The blood will be collected in streck tubes.
   For each tumor we will choose 2-4 breakpoints and see if these can be found in plasma from the same patients. A sensitive qPCR assay will be optimized for each mutation.
2. Plasma from all the 20 patients with HGSC will be analyzed for mutations in approximately 10 genes known to be mutated in HGSC. This is done using targeted sequencing, by using Illumina HiSeq 1500/ Tru seq tumor panel.
   Chromosomal copy number aberrations will be identified using whole genome sequencing.

Description of the cohort

The study is planned to include tumor tissue and blood samples from 20 patients. It is part of an undergraduate research year and has been designed so that it is possible to finish within 1 year, to see if the method in use is feasible. This will allow us to assess the perspectives of future research using liquid biopsies in HGSC. The Department of Gynecology and Obstetrics, OUH is treating all patients from the Region of Southern Denmark and a fraction of the patients from Region Zeeland with ovarian cancer. This means that it should be possible to recruit 20 women with HGSC for the project. Patients are included after information from a gynecologic specialist and signed consent.

People included in this study are patients with newly diagnosed high grade serous ovarian carcinomas.

Inclusion is taking place at the outpatient clinic at the Department of Gynecology and Obstetrics at Odense University Hospital. 

Data and biological material

From each patient a tumor sample and blood sample is collected and kept in a small biobank for Gynecologic cancers.

Furthermore information on the pathology of the tumor and tumor stage is used in the study.

Collaborating researchers and departments

Department of Clinical Genetics, Odense University Hospital

• Professor and Head of Research Torben A. Kruse
• Associate Professor Mads Thomassen
• PhD-student Kristina Magaard Koldby